537
JACC Vol. 17, No.2 February 1991 :537-42
DIAGNOSTIC STUDIES
Transesophageal Echocardiography Is Superior to Transthoracic Echocardiography in the Diagnosis of Sinus Venosus Atrial Septal Defect ITZHAK KRONZON, MD, FACC, PAUL A. TUNICK, MD, FACC, ROBIN S. FREEDBERG, MD, FACC, NARESH TREHAN, MD,* BARRY P. ROSENZWEIG, MD, FACC, MATTHEW E. SCHWINGER, MD New York, New York and New Delhi, India
The purpose of this study was to compare transthoracic and transesophageal echocardiography in the diagnosis of various types of atrial septal defects. Forty-one adult patients with the clinical diagnosis of atrial septal defect were studied by transthoracic and transesophageal echocardiography (30 women, 11 men; 18 to 81 years of age). Transthoracic echocardiography demonstrated the atrial septal defect in 33 patients (secundum type in 28, primum type in 3 and sinus venosus type in 2). Transesophageal echocardiography demonstrated the defect in aU 41 patients. Thus, in 8 (20%) of 41 patients the atrial septal defect was demonstrated by transesophageal and not by transthoracic echocardiography. Six of the eight had a sinus venosus type atrial septal defect; the other two patients had a secundum atrial septal defect (one of these two had a
technically poor transthoracic echocardiogram and the other had a small atrial septal defect). Transthoracic echocardiography, therefore, failed to demonstrate the sinus venosus defect in six (75 %) of eight patients. An anomalous venous connection associated with the sinus venosus defect was visualized by transesophageal echocardiography in seven of the eight patients but was not seen on transthoracic echocardiography in any patient. Sinus venosus type atrial septal defects are frequently not visualized in adults by conventional transthoracic echocardiography. Transesophageal echocardiography is recommended when an atrial septal defect is clinically suspected but cannot be visualized by transthoracic echocardiography. (J Am Coli CardioI1991;17:537-42)
Atrial septal defect is one of the most common congenital heart defects seen in adult cardiology practice. Its presence is frequently suggested by characteristic physical findings and is supported by typical radiographic and electrocardiographic features. In recent years. it has been diagnosed by echocardiography with the characteristic M-mode echocardiographic findings of right ventricular dilation and abnormal septal motion (right ventricular volume overload pattern) and actual demonstration of the defect by two-dimensional echocardiography (1). More recently, Doppler echocardiography has enabled the clinician to demonstrate the shunt direction, determine its magnitude and velocity (2,3) and evaluate pulmonary artery pressure by measuring the velocity of the tricuspid or pulmonary regurgitant jet (4). Although transthoracic echocardiography frequently visualizes ostium primum and ostium secundum defects, it has been reported (1) to be less sensitive in the demonstration of sinus venosus atrial septal defects. In recent years, trans-
esophageal echocardiography has been used to provide an alternative window for high resolution evaluation of abnormalities of the interatrial septum (5,6). In this study, we compared the value of transthoracic and transesophageal echocardiography in the assessment of various types ofatrial septal defect in adults.
From the Department of Medicine, New York University Medical Center. New York, New York and 'Department of Surgery. Escort Heart Institute, New Delhi. India. Manuscript received March 30,1990; revised manuscript received July 25. 1990, accepted August 3, 1990. Address for reprints: Itzhak Kronzon, MD, 560 First Avenue, Suite 2E, New York, New York 10016. ©1991 by the American College of Cardiology
Methods Study patients. Between May 1, 1988 and December 31, 1989, 41 patients with uncomplicated atrial septal defect were studied in the adult echocardiography laboratories of New York University Medical Center (Tisch Hospital) and Bellevue Hospital, New York and Escort Heart Institute and Research Center, New Delhi, India. The patients ranged in age from 18 to 81 years (mean 42 ± 15); there were 30 women and 11 men. Echocardiography. Each patient underwent routine transthoracic echocardiography that included M-mode and twodimensional imaging, pulsed and continuous wave Doppler ultrasound and color flow mapping. The interatrial septum was explored in the short-axis, apical four chamber and subxiphoid views. All patients also underwent transesophageal echocardiography performed with use of local anesthesia (by xylocaine spray) and no premedication. A 5 MHz 0735-1097/91/$3.50
538
KRONZON ET AL. SINUS VENOSUS ATRIAL SEPTAL DEFECT
Hewlett-Packard transesophageal echocardiography transducer was inserted through the mouth and advanced to a depth of 30 to 40 em. Images of the heart and great vessels were obtained and recorded on videotape for further analysis. There were no complications. The diagnosis of atrial septal defect was considered certain when the characteristic flow velocity pattern across the septum was identified (7). Attempts to identify anomalous pulmonary veins were made in each patient. Cardiac catheterization. This was performed in 28 of the 41 patients. The atrial septal defect was documented by oximetry and early hydrogen arrival time in the right atrium. The possibility of anomalous pulmonary venous return was evaluated by a method previously suggested by Danilowicz and Kronzon (8). With this method, the catheter is manipulated to a pulmonary vein and a contrast echocardiographic study is performed by injecting agitated saline solution. In patients with partial anomalous pulmonary venous return (and no right to left atrial shunt), the contrast agent appears mainly in the right-sided chambers, whereas in patients with normal pulmonary venous connection and an atrial septal defect, it appears first in the left atrium and then is shunted into the right chambers.
Results Transthoracic M·mode and two-dimensional echocardiography. Thirty-one of the 41 studies were considered excellent or good, 8 were considered fair and 2 poor. Right ventricular dilation was observed in 38 of the 41 patients and abnormal atrial septal motion in 33. Good subxiphoid views of the atrial septum were obtained in 31 patients and the atrial septal defect could be clearly seen in that view in 25. Doppler echocardiography demonstrated the characteristic left to right shunt in all 25 patients whose defect was clearly visualized in the subxiphoid view. In an additional eight patients without good demonstration of the defect in the subxiphoid view; the characteristic left to right shunt jet could clearly be demonstrated by pulsed, continuous wave and Doppler color ultrasound studies in another view, and the origin of the jet suggested the location of the defect in the interatrial septum. Thus, transthoracic echocardiography was considered diagnostic of atrial septal defect in 33 (80%) of 41 patients. Of these 33 defects, 3 were considered a primum, 28 a secundum and 2 a sinus venosus type defect. Therefore, there were eight patients in whom no characteristic shunt jet could be depicted by transthoracic echocardiography. The quality of the echocardiograms was considered good or excellent in six of these eight patients, fair in one and poor in one. Subxiphoid visualization of the interatrial septum was obtained in seven of the eight patients. Right ventricular, right atrial and pulmonary artery dilation were obvious in seven of the eight patients.
IACC Vol. 17. No.2
February 1991 :537-42
Transesophageal echocardiography. Transesophageal echocardiography demonstrated the atrial septum in all 41 patients studied; all studies were considered good to excellent, and the atrial septal defect was clearly seen in each patient (Fig. 1). Three patients had a primum, 30 a secundum and 8 a sinus venosus type atrial septal defect. In each patient, the shunt jet arising from the visualized defect could readily be demonstrated by Doppler color flow mapping (Fig. 2A). The timing, pattern and velocity of the jet could be further analyzed by M-mode color imaging and pulsed Doppler ultrasound (Fig. 2B). The atrial septal defect was visualized in all 41 patients, including the 8 patients in whom it was not demonstrated on routine transthoracic echocardiography. Of the eight defects visualized by transesophageal echocardiography and not by transthoracic echocardiography, six were in the most superior and posterior part of the atrial septum (that is, sinus venosus type atrial septal defect). In the other two patients, the defect was an ostium secundum type defect. One of these defects was quite small, and the other was observed in an obese patient with a very poor transthoracic echocardiographic study and no subxiphoid window. The comparison of transesophageal and transthoracic echocardiography in the diagnosis of atrial septal defect is summarized in Table 1. Of the eight adult patients with a sinus venosus atrial septal defect, the defect was visualized by transthoracic echocardiography in only two and could not be visualized in six (75%), but it was visualized by transesophageal echocardiography in all. The echocardiographic, cardiac catheterization and surgical findings in these eight patients are summarized in Table 2. Echocardiographic evaluation of partial anomalous pulmonary venous connection (Fig. 2C, 2D and 3). Despite many attempts, a clear diagnostic demonstration of partial anomalous pulmonary venous return could not be obtained by transthoracic echocardiography in any patient. Conversely, transesophageal echocardiography demonstrated an anomalous venous connection in seven of the eight patients with a sinus venosus type atrial septal defect. After the atrial septum was clearly visualized and the septal defect demonstrated, the transducer was slowly and carefully withdrawn to the junction between the superior vena cava and the right atrium. At this point, the tip was rotated or tilted, or both, until the entry of the anomalous pulmonary vein was demonstrated (Fig. 2C and 3A). The visualized anomalous vein entered the superior vena cava at its junction with the right atrium in four patients and entered the anterior aspect of the superior portion of the right atrium in one (Fig. 3B). In two patients, two separate anomalous veins were observed; one entered the superior vena cava and the other the superoanterior portion of the right atrium. Cardiac catheterization. Among the 30 patients who underwent cardiac catheterization, echocardiographic studies suggested that 3 had an ostium primum, 23 an ostium secundum and 4 a sinus venosus type defect. In each of the
KRONZON ET AL. SINUS Vl:.NOSUS ATRIAL SEPTAL DEFECT
JACC Vol. 17, No.2 February 1991:537-42
539
_fIiIiiII B
A
Figure 1. Transesophageal echocardiography in Ihe different types of atrial septal defect. Arrows point to the septal defect. A, Ostium primum defect at the most distal part of the septum; the proximal (superior) part is present. B, Ostium secundum mid-septal defect. C, Sinus venosus defect at the most proximal (superior) part of the septum. LA = left atrium; RA = right atrium; RV = right ventricle.
c 22 patients studied with hydrogen arrival curves from a pulmonary vein, the diagnosis of atrial septal defect was confirmed by early hydrogen arrival in the right atrium. Oximetry confirmed the presence of a left to right shunt in all patients with pulmonary to systemic flow ratios that ranged from 5:1 to 1.4:1. Echocardiographic findings of an atrial septal defect were confirmed in all 30 patients who underwent cardiac catheterization. Injection of agitated saline solution into an upper right pulmonary vein was performed in 14 patients. This injection confirmed the presence of an anomalous pulmonary vein in all four patients with a sinus venosus atrial septal defect studied by cardiac catheterization. The right upper pulmonary vein was normally connected to the left atrium in 9 of 10 patients with a secundum atrial septal defect; in 1
patient, the contrast echocardiographic study was not considered diagnostic because of poor contrast visualization within the heart. Operative findings. Twenty-eight of the 41 patients underwent surgery; 2 had an ostium primum defect, 21 an ostium secundum defect and 5 a sinus venosus atrial septal defect (that in 3 could not be demonstrated by transthoracic echocardiography). In each case, the location of the defect was as predicted by echocardiography. The presence and location of an anomalous pulmonary vein entering at or near the junction of the superior vena cava and the right atrium were confirmed in all five patients; in one patient, surgery revealed two anomalous veins whereas transesophageal echocardiography had demonstrated the drainage of only one.
540
KRONZON ET AL. SINUS VENOSUS ATRIAL SEPTAL DEFECT
JACC Vol. 17, No.2
February 1991 :537-42
A
B Figure 2. Doppler color flow mapping with esophageal echocardiography in sinus venosus type atrial septal defects. Transthoracic echocardiography did not disclose these defects. A, Left to right shunt jet (arrow) across the defect. B, Pulsed Doppler spectral tracing of the jet reveals the characteristic low velocity (I m/s) late systolic-holodiastolic flow pattern across a large atrial septal defect. C, The superior right pulmonary vein (arrow) draining into the superior vena cava (SVC). D, Aleft to right shunt jet is noted across a sinus venosus atrial septal defect (short arrow). This turbulent flow is away from the transducer, and appears as blue-green color. In addition, note the flow toward the transducer (red) in the right pulmonary vein (RPV), which is connected to the right atrium (RA). Other abbreviations as in Figure I.
Discussion
considered diagnostic only in this view. In the short-axis and four chamber views, atrial septal "dropout" may mimic an atrial septal defect. However, with the addition of Doppler ultrasound (especially color flow mapping), a characteristic shunt jet may be depicted. In uncomplicated atrial septal defects, this shunt jet is low velocity, left to right and continues from mid-systole to end-diastole (Fig. 2B). In eight patients in this study, the presence of such a jet originating from the atrial septum on Doppler echocardiography was Table 1. Comparison of Transthoracic (TIE) and Transesophageal Echocardiography (TEE) in the Demonstration of Atrial Septal Defect (ASD) in 41 Patients
Echocardiographic diagnosis of atrial septal defect.
M-mode and two-dimensional echocardiography are useful in the demonstration of atrial septal defects. Because the atrial septum is perpendicular to the ultrasound beam only in the subxiphoid view, the demonstration of the defect is
Primum ASD (n = 3) Secundum ASD (n = 30) Sinus venosus ASD (n = 8) Total
TIE
TEE
3 (100%)
3 (100%) 30 (100%) 8 (100%) 41 (100%)
28 (94%)
2 (25%) 33 (80%)
541
KRONZON ET AL. SINUS VENOSUS ATRIAL SEPTAL DEFECT
JACC Vol. 17. No.2 February \99\ :537-42
Table 2. Clinical Findings in Eight Patients With Sinus Venosus Atrial Septal Defect TEE
TTE PI. No.
3 4
5 6 7 8
Surgery
Cath
Age (yrll Gender
Quality of Study
Ox SVASO
Ox APVO
Ox SVASO
Ox APVO (no. of veins)
43/M 63/F 24/F 261M 46/F 23/F 56/F 31/F
Good Good Excellent Good Excellent Fair Excellent Good
No No Yes No Yes No No No
No No No No No No No No
Yes Yes Yes Yes Yes Yes Yes Yes
No Yes (I) Yes (I) Yes (2) Yes (1) Yes (1) Yes (2) Yes (I)
Ox ASO
Ox APVO
Ox ASO
Ox APVO (no. of veins)
Yes
Yes
Yes Yes
Yes (2) Yes (I)
Yes
Yes
Yes
Yes (I)
Yes Yes
Yes Yes
Yes Yes
Yes (2) Yes (I)
Cath = cardiac catheterization; Ox APVD = demonstration of anomalous pulmonary venous drainage: Ox ASO = demonstration of atrial septal defect: Dx SVASD = demonstration of sinus venosus atrial septal defect: F = female: M = male; PI. = patient: TEE = transesophageal echocardiography; TTE = transthoracic echocardiography: ... = not done.
considered diagnostic of atrial septal defect even when diagnostic subxiphoid views could not be obtained. Transthoracic versus transesophageal imaging. In 154 adults and children who underwent transthoracic echocardiography, Shub et al. (I) demonstrated 100% of ostium primum and >90% of ostium secundum atrial septal defects but only 44% of sinus venosus type defects. Because the Figure 3. Transesophageal echocardiography demonstrating anomalous pulmonary venous drainage in two patients with a sinus venosus type atrial septal defect. Transthoracic imaging revealed neither these defects nor the anomalous pulmonary venous drainage. A, The right pulmonary vein (RPV) drains into the most distal segment of the superior vena cava (SVC) near its junction with the right atrium. The arrow shows the connection site of an additional vein joining the abnormal drainage. B, The site of the sinus venosus defect is noted by the arrows. An anomalous right pulmonary vein drains into the right atrium (RA). AO = aorta: other abbreviations as in Figures I and 2.
A
B
latter type occurs in up to 20% of adults with an atrial septal defect, the insensitivity of the transthoracic echocardiographic approach in diagnosing it is a significant limitation. The failure to visualize this type of defect is probably a result of its location in an echocardiographic blind spot. Indeed, the junction of the superior vena cava with the right atrium as well as the most proximal portion of the septum and most superior part of the right atrium are frequently not visualized by the transthoracic approach in the adult. Anomalous pulmonary venous return. Partial anomalous pulmonary venous return from the upper (and sometimes middle) right pulmonary lobes is almost always associated with a sinus venosus type atrial septal defect rather than with other types of defect. However. an anomalous vein from the superior or middle right pulmonary lobes usually cannot be visualized by transthoracic echocardiography. The superior vena cava may be visualized by the suprasternal approach
542
KRONZON ET AL. SINUS VENOSUS ATRIAL SEPTAL DEFECT
(9); however, the anomalous pulmonary veins in our patients could not be seen with use of this view. Previous studies. Transesophageal echocardiography has been shown to demonstrate the atrial septum with high resolution. The anatomy of the interatrial septum can be analyzed in detail. Hanrath et al. (6) showed that transesophageal echocardiography is superior to transthoracic echocardiography in the demonstration of ostium secundum atrial septal defect. Weigel et al. (10) observed that transesophageal echocardiography was always diagnostic in a group of patients with various types of atrial septal defect and poor transthoracic echocardiograms. A recent report from our laboratory (II) demonstrated that even very small defects, such as those created by transseptal puncture during percutaneous balloon valvuloplasty, can be demonstrated by transesophageal echocardiography much more readily than by transthoracic echocardiography. Oh et al. (12) described a patient with a sinus venosus type atrial septal defect in whom transthoracic echocardiography failed to demonstrate the defect and transesophageal echocardiography clearly showed it. Sinus venosus atrial septal defect. In our series, transthoracic echocardiography did not demonstrate a sinus venosus atrial septal defect in six of the eight patients who had the defect. It did not demonstrate the partial anomalous pulmonary venous connection in any of the patients studied. In contrast, transesophageal echocardiography clearly showed the defect in all eight patients and showed an anomalous venous connection in seven of the eight. Transesophageal echocardiography appears to provide better information in patients with a sinus venosus atrial septal defect because of its higher resolution and the better exposure of the most proximal portion of the atrial septum and its related structures (13). Limitations. In this study, the noninvasive reference standard for the presence of an atrial septal defect was the demonstration of the characteristic flow jet across a visualized defect or emanating from the interatri;ll septum. It appears that transesophageal echocardiography has a higher sensitivity than transthoracic echocardiography in detecting the defect (especially a sinus venosus defect). Although the transesophageal echocardiographic findings were confirmed by cardiac catheterization in all patients studied and by surgery in all patients operated on, its specificity is unknown. The finding of anomalous pulmonary venous return was confirmed invasively or surgically only in a small group of patients, and it is not known if all anomalous veins can be detected by transesophageal echocardiography. In this study, three of the eight patients with an atrial septal defect diagnosed by transesophageal echocardiography alone had neither cardiac catheterization nor surgical intervention. It
JACC Vol. 17. No.2 February 1991 :537-42
should also be noted that Doppler color flow velocities measure flow velocity and not flow, and therefore color flow mapping may lack sensitivity in a low flow atrial septal defect. Transesophageal echocardiography is an invasive test, and rare complications have been reported (14). It cannot be recommended in the routine evaluation of a patient who is referred to the echocardiographic laboratory. Clinical implications. Whenever the clinical findings or results of transthoracic echocardiography, or both, are suggestive but not diagnostic of atrial septal defect, a sinus venosus type defect should be considered and transesophageal echocardiography performed for visualization of the defect and an associated pulmonary venous connection.
References I. Shub C. Dimopoulos IN, Seward 18. Sensitivity of two dimensional echocardiography in the direct visualization of atrial septal defect utilizing the subcostal approach: experience with 154 patients. J Am Coli Cardiol 1983:2: 127-35. 2. Hatle L. Angelsen B. Doppler Ultrasound in Cardiology. 2nd ed. Philadelphia: Lea & Febiger. 1985:228-36. 3. Suzuki Y. Kambara H. Kadota K, et al. Detection of intracardiac shunt flow in atrial septal defect using a real time, two-dimensional, color-coded Doppler flow imaging system and comparison with contrast two dimensional echocardiography. Am J Cardiol 1985;56:347-50. 4. Yock PG, Popp RL. Noninvasive estimation of right ventricular systolic pressure by Doppler ultrasound in patients with tricuspid regurgitation. Circulation 1984:70:657-62. 5. Schwinger ME, Gindea AJ. Kronzon I. The anatomy of the interatrial septum: a transesophageal echocardiographic study. Am Heart 1 1990: 119: 1401-5. 6. Hanrath P. Schlutter M. Langenstein BA, et al. Detection of ostium secundum atrial septal defects by transesophageal cross sectional echocardiography. Br Heart 1 1983:49:350-8. 7. Hatle L, Angelsen B. Pulsed and continuous wave Doppler in diagnosis and assessment of various heart lesions. In Ref 2:228-32. 8. Danilowicz D. Kronzon I. Contrast echocardiography in the diagnosis of partial anomalous pulmonary venous connection. Am J Cardiol 1979:43: 248-52. 9. Cohen ML. Cohen B. Kronzon I. Lighty GW lr, Politzer F, Winer HE. Superior vena caval blood flow velocities in normal adults: a Doppler echocardiographic study. 1 Appl Physiol 1986;61 :215-9. 10. Weigel JT, Seward 18, Hagler Dl. Khanderia BK, Tajik lA. Transesophageal echocardiography in 21 atrial septal defects with incomplete precordial echocardiography (abstr). Circulation 1989:80(supplll):1l-474. II. Kronzon I. Tunick PA. Goldfarb A, et al. Echocardiographic and hemodynamic characteristics of atrial septal defects created by percutaneous valvuloplasty. 1 Am Soc Echo 1990;3:64-71. 12. Oh lK, Seward lB. Khanderia BK, Danielson GK. Tajik AJ. Visualization of sinus venosus atrial septal defect by transesophageal echocardiography. 1 Am Soc Echo 1988:1:275-7. 13. Weigel TJ. Seward JB. Hagler DJ. Transesophageal echocardiography in , anomalous pulmonary and systemic venous connection (abstr). Circulation 1989:80(suppI1l):1I-339. 14. Daniel WG, Mugge A, Schroder E, Wunzlaft P. Grote J. Transesophageal echocardiography in clinical cardiology: indications, practicability and risk (abstr). Circulation 1988:78(supplll):1l-296.
JACC Vol. 17, No.2 February 1991 :537-42
DIAGNOSTIC STUDIES
Transesophageal Echocardiography Is Superior to Transthoracic Echocardiography in the Diagnosis of Sinus Venosus Atrial Septal Defect ITZHAK KRONZON, MD, FACC, PAUL A. TUNICK, MD, FACC, ROBIN S. FREEDBERG, MD, FACC, NARESH TREHAN, MD,* BARRY P. ROSENZWEIG, MD, FACC, MATTHEW E. SCHWINGER, MD New York, New York and New Delhi, India
The purpose of this study was to compare transthoracic and transesophageal echocardiography in the diagnosis of various types of atrial septal defects. Forty-one adult patients with the clinical diagnosis of atrial septal defect were studied by transthoracic and transesophageal echocardiography (30 women, 11 men; 18 to 81 years of age). Transthoracic echocardiography demonstrated the atrial septal defect in 33 patients (secundum type in 28, primum type in 3 and sinus venosus type in 2). Transesophageal echocardiography demonstrated the defect in aU 41 patients. Thus, in 8 (20%) of 41 patients the atrial septal defect was demonstrated by transesophageal and not by transthoracic echocardiography. Six of the eight had a sinus venosus type atrial septal defect; the other two patients had a secundum atrial septal defect (one of these two had a
technically poor transthoracic echocardiogram and the other had a small atrial septal defect). Transthoracic echocardiography, therefore, failed to demonstrate the sinus venosus defect in six (75 %) of eight patients. An anomalous venous connection associated with the sinus venosus defect was visualized by transesophageal echocardiography in seven of the eight patients but was not seen on transthoracic echocardiography in any patient. Sinus venosus type atrial septal defects are frequently not visualized in adults by conventional transthoracic echocardiography. Transesophageal echocardiography is recommended when an atrial septal defect is clinically suspected but cannot be visualized by transthoracic echocardiography. (J Am Coli CardioI1991;17:537-42)
Atrial septal defect is one of the most common congenital heart defects seen in adult cardiology practice. Its presence is frequently suggested by characteristic physical findings and is supported by typical radiographic and electrocardiographic features. In recent years. it has been diagnosed by echocardiography with the characteristic M-mode echocardiographic findings of right ventricular dilation and abnormal septal motion (right ventricular volume overload pattern) and actual demonstration of the defect by two-dimensional echocardiography (1). More recently, Doppler echocardiography has enabled the clinician to demonstrate the shunt direction, determine its magnitude and velocity (2,3) and evaluate pulmonary artery pressure by measuring the velocity of the tricuspid or pulmonary regurgitant jet (4). Although transthoracic echocardiography frequently visualizes ostium primum and ostium secundum defects, it has been reported (1) to be less sensitive in the demonstration of sinus venosus atrial septal defects. In recent years, trans-
esophageal echocardiography has been used to provide an alternative window for high resolution evaluation of abnormalities of the interatrial septum (5,6). In this study, we compared the value of transthoracic and transesophageal echocardiography in the assessment of various types ofatrial septal defect in adults.
From the Department of Medicine, New York University Medical Center. New York, New York and 'Department of Surgery. Escort Heart Institute, New Delhi. India. Manuscript received March 30,1990; revised manuscript received July 25. 1990, accepted August 3, 1990. Address for reprints: Itzhak Kronzon, MD, 560 First Avenue, Suite 2E, New York, New York 10016. ©1991 by the American College of Cardiology
Methods Study patients. Between May 1, 1988 and December 31, 1989, 41 patients with uncomplicated atrial septal defect were studied in the adult echocardiography laboratories of New York University Medical Center (Tisch Hospital) and Bellevue Hospital, New York and Escort Heart Institute and Research Center, New Delhi, India. The patients ranged in age from 18 to 81 years (mean 42 ± 15); there were 30 women and 11 men. Echocardiography. Each patient underwent routine transthoracic echocardiography that included M-mode and twodimensional imaging, pulsed and continuous wave Doppler ultrasound and color flow mapping. The interatrial septum was explored in the short-axis, apical four chamber and subxiphoid views. All patients also underwent transesophageal echocardiography performed with use of local anesthesia (by xylocaine spray) and no premedication. A 5 MHz 0735-1097/91/$3.50
538
KRONZON ET AL. SINUS VENOSUS ATRIAL SEPTAL DEFECT
Hewlett-Packard transesophageal echocardiography transducer was inserted through the mouth and advanced to a depth of 30 to 40 em. Images of the heart and great vessels were obtained and recorded on videotape for further analysis. There were no complications. The diagnosis of atrial septal defect was considered certain when the characteristic flow velocity pattern across the septum was identified (7). Attempts to identify anomalous pulmonary veins were made in each patient. Cardiac catheterization. This was performed in 28 of the 41 patients. The atrial septal defect was documented by oximetry and early hydrogen arrival time in the right atrium. The possibility of anomalous pulmonary venous return was evaluated by a method previously suggested by Danilowicz and Kronzon (8). With this method, the catheter is manipulated to a pulmonary vein and a contrast echocardiographic study is performed by injecting agitated saline solution. In patients with partial anomalous pulmonary venous return (and no right to left atrial shunt), the contrast agent appears mainly in the right-sided chambers, whereas in patients with normal pulmonary venous connection and an atrial septal defect, it appears first in the left atrium and then is shunted into the right chambers.
Results Transthoracic M·mode and two-dimensional echocardiography. Thirty-one of the 41 studies were considered excellent or good, 8 were considered fair and 2 poor. Right ventricular dilation was observed in 38 of the 41 patients and abnormal atrial septal motion in 33. Good subxiphoid views of the atrial septum were obtained in 31 patients and the atrial septal defect could be clearly seen in that view in 25. Doppler echocardiography demonstrated the characteristic left to right shunt in all 25 patients whose defect was clearly visualized in the subxiphoid view. In an additional eight patients without good demonstration of the defect in the subxiphoid view; the characteristic left to right shunt jet could clearly be demonstrated by pulsed, continuous wave and Doppler color ultrasound studies in another view, and the origin of the jet suggested the location of the defect in the interatrial septum. Thus, transthoracic echocardiography was considered diagnostic of atrial septal defect in 33 (80%) of 41 patients. Of these 33 defects, 3 were considered a primum, 28 a secundum and 2 a sinus venosus type defect. Therefore, there were eight patients in whom no characteristic shunt jet could be depicted by transthoracic echocardiography. The quality of the echocardiograms was considered good or excellent in six of these eight patients, fair in one and poor in one. Subxiphoid visualization of the interatrial septum was obtained in seven of the eight patients. Right ventricular, right atrial and pulmonary artery dilation were obvious in seven of the eight patients.
IACC Vol. 17. No.2
February 1991 :537-42
Transesophageal echocardiography. Transesophageal echocardiography demonstrated the atrial septum in all 41 patients studied; all studies were considered good to excellent, and the atrial septal defect was clearly seen in each patient (Fig. 1). Three patients had a primum, 30 a secundum and 8 a sinus venosus type atrial septal defect. In each patient, the shunt jet arising from the visualized defect could readily be demonstrated by Doppler color flow mapping (Fig. 2A). The timing, pattern and velocity of the jet could be further analyzed by M-mode color imaging and pulsed Doppler ultrasound (Fig. 2B). The atrial septal defect was visualized in all 41 patients, including the 8 patients in whom it was not demonstrated on routine transthoracic echocardiography. Of the eight defects visualized by transesophageal echocardiography and not by transthoracic echocardiography, six were in the most superior and posterior part of the atrial septum (that is, sinus venosus type atrial septal defect). In the other two patients, the defect was an ostium secundum type defect. One of these defects was quite small, and the other was observed in an obese patient with a very poor transthoracic echocardiographic study and no subxiphoid window. The comparison of transesophageal and transthoracic echocardiography in the diagnosis of atrial septal defect is summarized in Table 1. Of the eight adult patients with a sinus venosus atrial septal defect, the defect was visualized by transthoracic echocardiography in only two and could not be visualized in six (75%), but it was visualized by transesophageal echocardiography in all. The echocardiographic, cardiac catheterization and surgical findings in these eight patients are summarized in Table 2. Echocardiographic evaluation of partial anomalous pulmonary venous connection (Fig. 2C, 2D and 3). Despite many attempts, a clear diagnostic demonstration of partial anomalous pulmonary venous return could not be obtained by transthoracic echocardiography in any patient. Conversely, transesophageal echocardiography demonstrated an anomalous venous connection in seven of the eight patients with a sinus venosus type atrial septal defect. After the atrial septum was clearly visualized and the septal defect demonstrated, the transducer was slowly and carefully withdrawn to the junction between the superior vena cava and the right atrium. At this point, the tip was rotated or tilted, or both, until the entry of the anomalous pulmonary vein was demonstrated (Fig. 2C and 3A). The visualized anomalous vein entered the superior vena cava at its junction with the right atrium in four patients and entered the anterior aspect of the superior portion of the right atrium in one (Fig. 3B). In two patients, two separate anomalous veins were observed; one entered the superior vena cava and the other the superoanterior portion of the right atrium. Cardiac catheterization. Among the 30 patients who underwent cardiac catheterization, echocardiographic studies suggested that 3 had an ostium primum, 23 an ostium secundum and 4 a sinus venosus type defect. In each of the
KRONZON ET AL. SINUS Vl:.NOSUS ATRIAL SEPTAL DEFECT
JACC Vol. 17, No.2 February 1991:537-42
539
_fIiIiiII B
A
Figure 1. Transesophageal echocardiography in Ihe different types of atrial septal defect. Arrows point to the septal defect. A, Ostium primum defect at the most distal part of the septum; the proximal (superior) part is present. B, Ostium secundum mid-septal defect. C, Sinus venosus defect at the most proximal (superior) part of the septum. LA = left atrium; RA = right atrium; RV = right ventricle.
c 22 patients studied with hydrogen arrival curves from a pulmonary vein, the diagnosis of atrial septal defect was confirmed by early hydrogen arrival in the right atrium. Oximetry confirmed the presence of a left to right shunt in all patients with pulmonary to systemic flow ratios that ranged from 5:1 to 1.4:1. Echocardiographic findings of an atrial septal defect were confirmed in all 30 patients who underwent cardiac catheterization. Injection of agitated saline solution into an upper right pulmonary vein was performed in 14 patients. This injection confirmed the presence of an anomalous pulmonary vein in all four patients with a sinus venosus atrial septal defect studied by cardiac catheterization. The right upper pulmonary vein was normally connected to the left atrium in 9 of 10 patients with a secundum atrial septal defect; in 1
patient, the contrast echocardiographic study was not considered diagnostic because of poor contrast visualization within the heart. Operative findings. Twenty-eight of the 41 patients underwent surgery; 2 had an ostium primum defect, 21 an ostium secundum defect and 5 a sinus venosus atrial septal defect (that in 3 could not be demonstrated by transthoracic echocardiography). In each case, the location of the defect was as predicted by echocardiography. The presence and location of an anomalous pulmonary vein entering at or near the junction of the superior vena cava and the right atrium were confirmed in all five patients; in one patient, surgery revealed two anomalous veins whereas transesophageal echocardiography had demonstrated the drainage of only one.
540
KRONZON ET AL. SINUS VENOSUS ATRIAL SEPTAL DEFECT
JACC Vol. 17, No.2
February 1991 :537-42
A
B Figure 2. Doppler color flow mapping with esophageal echocardiography in sinus venosus type atrial septal defects. Transthoracic echocardiography did not disclose these defects. A, Left to right shunt jet (arrow) across the defect. B, Pulsed Doppler spectral tracing of the jet reveals the characteristic low velocity (I m/s) late systolic-holodiastolic flow pattern across a large atrial septal defect. C, The superior right pulmonary vein (arrow) draining into the superior vena cava (SVC). D, Aleft to right shunt jet is noted across a sinus venosus atrial septal defect (short arrow). This turbulent flow is away from the transducer, and appears as blue-green color. In addition, note the flow toward the transducer (red) in the right pulmonary vein (RPV), which is connected to the right atrium (RA). Other abbreviations as in Figure I.
Discussion
considered diagnostic only in this view. In the short-axis and four chamber views, atrial septal "dropout" may mimic an atrial septal defect. However, with the addition of Doppler ultrasound (especially color flow mapping), a characteristic shunt jet may be depicted. In uncomplicated atrial septal defects, this shunt jet is low velocity, left to right and continues from mid-systole to end-diastole (Fig. 2B). In eight patients in this study, the presence of such a jet originating from the atrial septum on Doppler echocardiography was Table 1. Comparison of Transthoracic (TIE) and Transesophageal Echocardiography (TEE) in the Demonstration of Atrial Septal Defect (ASD) in 41 Patients
Echocardiographic diagnosis of atrial septal defect.
M-mode and two-dimensional echocardiography are useful in the demonstration of atrial septal defects. Because the atrial septum is perpendicular to the ultrasound beam only in the subxiphoid view, the demonstration of the defect is
Primum ASD (n = 3) Secundum ASD (n = 30) Sinus venosus ASD (n = 8) Total
TIE
TEE
3 (100%)
3 (100%) 30 (100%) 8 (100%) 41 (100%)
28 (94%)
2 (25%) 33 (80%)
541
KRONZON ET AL. SINUS VENOSUS ATRIAL SEPTAL DEFECT
JACC Vol. 17. No.2 February \99\ :537-42
Table 2. Clinical Findings in Eight Patients With Sinus Venosus Atrial Septal Defect TEE
TTE PI. No.
3 4
5 6 7 8
Surgery
Cath
Age (yrll Gender
Quality of Study
Ox SVASO
Ox APVO
Ox SVASO
Ox APVO (no. of veins)
43/M 63/F 24/F 261M 46/F 23/F 56/F 31/F
Good Good Excellent Good Excellent Fair Excellent Good
No No Yes No Yes No No No
No No No No No No No No
Yes Yes Yes Yes Yes Yes Yes Yes
No Yes (I) Yes (I) Yes (2) Yes (1) Yes (1) Yes (2) Yes (I)
Ox ASO
Ox APVO
Ox ASO
Ox APVO (no. of veins)
Yes
Yes
Yes Yes
Yes (2) Yes (I)
Yes
Yes
Yes
Yes (I)
Yes Yes
Yes Yes
Yes Yes
Yes (2) Yes (I)
Cath = cardiac catheterization; Ox APVD = demonstration of anomalous pulmonary venous drainage: Ox ASO = demonstration of atrial septal defect: Dx SVASD = demonstration of sinus venosus atrial septal defect: F = female: M = male; PI. = patient: TEE = transesophageal echocardiography; TTE = transthoracic echocardiography: ... = not done.
considered diagnostic of atrial septal defect even when diagnostic subxiphoid views could not be obtained. Transthoracic versus transesophageal imaging. In 154 adults and children who underwent transthoracic echocardiography, Shub et al. (I) demonstrated 100% of ostium primum and >90% of ostium secundum atrial septal defects but only 44% of sinus venosus type defects. Because the Figure 3. Transesophageal echocardiography demonstrating anomalous pulmonary venous drainage in two patients with a sinus venosus type atrial septal defect. Transthoracic imaging revealed neither these defects nor the anomalous pulmonary venous drainage. A, The right pulmonary vein (RPV) drains into the most distal segment of the superior vena cava (SVC) near its junction with the right atrium. The arrow shows the connection site of an additional vein joining the abnormal drainage. B, The site of the sinus venosus defect is noted by the arrows. An anomalous right pulmonary vein drains into the right atrium (RA). AO = aorta: other abbreviations as in Figures I and 2.
A
B
latter type occurs in up to 20% of adults with an atrial septal defect, the insensitivity of the transthoracic echocardiographic approach in diagnosing it is a significant limitation. The failure to visualize this type of defect is probably a result of its location in an echocardiographic blind spot. Indeed, the junction of the superior vena cava with the right atrium as well as the most proximal portion of the septum and most superior part of the right atrium are frequently not visualized by the transthoracic approach in the adult. Anomalous pulmonary venous return. Partial anomalous pulmonary venous return from the upper (and sometimes middle) right pulmonary lobes is almost always associated with a sinus venosus type atrial septal defect rather than with other types of defect. However. an anomalous vein from the superior or middle right pulmonary lobes usually cannot be visualized by transthoracic echocardiography. The superior vena cava may be visualized by the suprasternal approach
542
KRONZON ET AL. SINUS VENOSUS ATRIAL SEPTAL DEFECT
(9); however, the anomalous pulmonary veins in our patients could not be seen with use of this view. Previous studies. Transesophageal echocardiography has been shown to demonstrate the atrial septum with high resolution. The anatomy of the interatrial septum can be analyzed in detail. Hanrath et al. (6) showed that transesophageal echocardiography is superior to transthoracic echocardiography in the demonstration of ostium secundum atrial septal defect. Weigel et al. (10) observed that transesophageal echocardiography was always diagnostic in a group of patients with various types of atrial septal defect and poor transthoracic echocardiograms. A recent report from our laboratory (II) demonstrated that even very small defects, such as those created by transseptal puncture during percutaneous balloon valvuloplasty, can be demonstrated by transesophageal echocardiography much more readily than by transthoracic echocardiography. Oh et al. (12) described a patient with a sinus venosus type atrial septal defect in whom transthoracic echocardiography failed to demonstrate the defect and transesophageal echocardiography clearly showed it. Sinus venosus atrial septal defect. In our series, transthoracic echocardiography did not demonstrate a sinus venosus atrial septal defect in six of the eight patients who had the defect. It did not demonstrate the partial anomalous pulmonary venous connection in any of the patients studied. In contrast, transesophageal echocardiography clearly showed the defect in all eight patients and showed an anomalous venous connection in seven of the eight. Transesophageal echocardiography appears to provide better information in patients with a sinus venosus atrial septal defect because of its higher resolution and the better exposure of the most proximal portion of the atrial septum and its related structures (13). Limitations. In this study, the noninvasive reference standard for the presence of an atrial septal defect was the demonstration of the characteristic flow jet across a visualized defect or emanating from the interatri;ll septum. It appears that transesophageal echocardiography has a higher sensitivity than transthoracic echocardiography in detecting the defect (especially a sinus venosus defect). Although the transesophageal echocardiographic findings were confirmed by cardiac catheterization in all patients studied and by surgery in all patients operated on, its specificity is unknown. The finding of anomalous pulmonary venous return was confirmed invasively or surgically only in a small group of patients, and it is not known if all anomalous veins can be detected by transesophageal echocardiography. In this study, three of the eight patients with an atrial septal defect diagnosed by transesophageal echocardiography alone had neither cardiac catheterization nor surgical intervention. It
JACC Vol. 17. No.2 February 1991 :537-42
should also be noted that Doppler color flow velocities measure flow velocity and not flow, and therefore color flow mapping may lack sensitivity in a low flow atrial septal defect. Transesophageal echocardiography is an invasive test, and rare complications have been reported (14). It cannot be recommended in the routine evaluation of a patient who is referred to the echocardiographic laboratory. Clinical implications. Whenever the clinical findings or results of transthoracic echocardiography, or both, are suggestive but not diagnostic of atrial septal defect, a sinus venosus type defect should be considered and transesophageal echocardiography performed for visualization of the defect and an associated pulmonary venous connection.
References I. Shub C. Dimopoulos IN, Seward 18. Sensitivity of two dimensional echocardiography in the direct visualization of atrial septal defect utilizing the subcostal approach: experience with 154 patients. J Am Coli Cardiol 1983:2: 127-35. 2. Hatle L. Angelsen B. Doppler Ultrasound in Cardiology. 2nd ed. Philadelphia: Lea & Febiger. 1985:228-36. 3. Suzuki Y. Kambara H. Kadota K, et al. Detection of intracardiac shunt flow in atrial septal defect using a real time, two-dimensional, color-coded Doppler flow imaging system and comparison with contrast two dimensional echocardiography. Am J Cardiol 1985;56:347-50. 4. Yock PG, Popp RL. Noninvasive estimation of right ventricular systolic pressure by Doppler ultrasound in patients with tricuspid regurgitation. Circulation 1984:70:657-62. 5. Schwinger ME, Gindea AJ. Kronzon I. The anatomy of the interatrial septum: a transesophageal echocardiographic study. Am Heart 1 1990: 119: 1401-5. 6. Hanrath P. Schlutter M. Langenstein BA, et al. Detection of ostium secundum atrial septal defects by transesophageal cross sectional echocardiography. Br Heart 1 1983:49:350-8. 7. Hatle L, Angelsen B. Pulsed and continuous wave Doppler in diagnosis and assessment of various heart lesions. In Ref 2:228-32. 8. Danilowicz D. Kronzon I. Contrast echocardiography in the diagnosis of partial anomalous pulmonary venous connection. Am J Cardiol 1979:43: 248-52. 9. Cohen ML. Cohen B. Kronzon I. Lighty GW lr, Politzer F, Winer HE. Superior vena caval blood flow velocities in normal adults: a Doppler echocardiographic study. 1 Appl Physiol 1986;61 :215-9. 10. Weigel JT, Seward 18, Hagler Dl. Khanderia BK, Tajik lA. Transesophageal echocardiography in 21 atrial septal defects with incomplete precordial echocardiography (abstr). Circulation 1989:80(supplll):1l-474. II. Kronzon I. Tunick PA. Goldfarb A, et al. Echocardiographic and hemodynamic characteristics of atrial septal defects created by percutaneous valvuloplasty. 1 Am Soc Echo 1990;3:64-71. 12. Oh lK, Seward lB. Khanderia BK, Danielson GK. Tajik AJ. Visualization of sinus venosus atrial septal defect by transesophageal echocardiography. 1 Am Soc Echo 1988:1:275-7. 13. Weigel TJ. Seward JB. Hagler DJ. Transesophageal echocardiography in , anomalous pulmonary and systemic venous connection (abstr). Circulation 1989:80(suppI1l):1I-339. 14. Daniel WG, Mugge A, Schroder E, Wunzlaft P. Grote J. Transesophageal echocardiography in clinical cardiology: indications, practicability and risk (abstr). Circulation 1988:78(supplll):1l-296.
