- Email: [email protected]
Postoperative evaluation of pulmonary arteries in congenital heart surgery by magnetic resonance imaging: Comparison with echocardiography
Postopefative evaluation of pulmonary arteries in congenital heart surgery by magnetic resonance imaging: Comparison with echocardiography Palliative and corrective operations for the treatment of cyanotic congenital heart disease frequently involve or potentially influence the size of the pulmonary arteries. Echocardiography and magnetic resonance imaging (MRI) are two noninvasive imaging techniques currently used to assess morphologic abnormalities of the pulmonary arteries. The purpose of thk study was to evaluate the role of NW in comparison with echocardiography for deffning morphologic changes of the pulmonary arteries after congenital heart surgery. The YRI scans and echocardiograms of 33 patients with surgery involving or affecting the pulmonary arteries were compared. The pulmonary outflow tract, pulmonary confluence, right and left pulmonary arteries, and surgical shunts were separately evaluated. Cineangiography and surgical reports were used to confirm findings. MRI and echocardiography were equivalent for demonstmting abnormafiies of the right ventricular outflow tract, main pulmonary artery, and a variety of pulmonary shunts. MRI was superior to echocardiography in demonstrating abnormalities of the right and left pulmonary arterial branches (p < 0.001). MRI is effective for monitoring pulmonary arterial status after surgery and is superior to echocardiography for the evaluation of the right and left pulmonary arteries. (AM HEART J 1994;128:1139-46.)
Andre J. Duerinckx, MD, PhD, al b Lewis Wexler, MD,” Anirban Banerjee, MD,“, d Sarah S. Higgins, PhDF Christian E. Hardy, MD: Gregg Helton, MD,” John Rhee, MD,” Soroosh Mahboubi, MD,f and Charles B. Higgins, MDa San Francisco, Los Angeles, Stanford, Philadelphia, Pa.
and Oakland,
Palliative and corrective operations for the treatment of cyanotic congenital heart disease frequently involve or potentially influence the size of the pulmonary arteries. After surgery it is usually necessary to sequentially evaluate the pulmonary arteries to detect occult stenoses or to monitor growth of the pulmonary arteries in response to increased blood fl0w.l In patients with pulmonary atresia and venFrom the *Department of Radiology, Magnetic Resonance Imaging Section, University of California, San Francisco; the bRadiology Service, Magnetic Resonance Imaging, Veterans Administration Medical Center, Los Angeles; the CDepartment of Radiology, Stanford University; the dDivision of Cardiology, Children’s Hospital Medical Center, Cincinnati; ePediatric Cardiology, Children’s Hospital Oakland, and the department of Radiology, The Children’s Hospital of Philadelphia. Received
for publication
Jan. 20, 1994; accepted
March
7, 1994.
Reprint requests: Charles B. Higgins, MD, Department of Radiology, MRI Section, University of California, 505 Parnassus Ave., Suite L306, San Francisco, CA 94143-0628. Copyright @ 1994 by Mosby-Year 0002~8703/94/$3.00 + 0 4/l/58600
Book,
Inc.
Calif.,
Cincinnati,
Ohio, and
tricular septal defect, the size of the pulmonary arteries is a prime determinant in the appropriateness of a one-stage surgical correction or the need for preliminary palliative surgical procedures to stimulate pulmonary arterial growth that is compatible with later repair. After initial palliative surgery, evaluation of interval growth of pulmonary arteries is also needed before a method of complete correction is chosen. The techniques currently used to assess the pulmonary arteries are angiocardiography, echocardiography (echo), and magnetic resonance imaging (MRI). Angiography has limitations in visualizing pulmonary arteries in patients with pulmonary atresia and predominate flow to the lungs by aortopulmonary collaterals. Echocardiography is sometimes limited in demonstrating the pulmonary arteries by poor sound transmission because of overlying lung, especially in older children. MRI is another noninvasive method for evaluating congenital heart disease, in1139
1140
December 1994
Duerinckx et al.
eluding palliative systemic pulmonary artery shunts, central and proximal pulmonary arteries, and greatvessel abnormalities.2-8 MRI has been shown to be effective in the evaluation of patients with obstruction of the right ventricular outflow tract3 and in patients with pulmonary atresia.2T 7 Although previous studies2-7 have shown the effectiveness of MRI in evaluating the pulmonary arteries, only a single study3 shows comparisons between MRI and echo for the evaluation of pulmonary arterial abnormalities. No studies have provided comparative data specifically in patients after surgery involving the pulmonary arteries. The purpose of the present study was to compare MRI to echo in the evaluation of the morphologic changes of the pulmonary arteries in patients with congenital heart disease after surgery that might influence the size and/or configuration of the pulmonary arteries. Results of the noninvasive modalities were confirmed from angiography and/or surgical findings. METHODS Study population. Seventy-two patients with congenital heart disease in whom prior surgery had involved the pulmonary arteries were referred for MRI from 1988 to 1992. Of the original 72 patients, 33 had echocardiograms performed within 12 months of the MRI study that were available for review, and no intervening surgery or treatment had occurred between the two studies. The ages of the 33 patients ranged from 1 month to 26 years (mean 4.5 years). Patients were excluded from the study if (1) there was intervening surgery between the dates of the MRI and echocardiography studies; (2) one of the studies was missing or not available for review; or (3) there was a >12-month interval between the two studies. The average time difference between the time of the MRI and echocardiography was 2.5 months (range 0 days to 12 months). Major cardiovascular anomalies in this group of 33 patients were varied: 8 had tetralogy of Fallot, 9 transposition of the great arteries, 2 pulmonary atresia with intact ventricular septum, 3 single ventricle, 2 hypoplastic left heart syndrome, 3 truncus arteriosus, 1 pulmonic stenosis, 2 double-outlet right ventricle, 1 Ebstein’s anomaly, 1 tricuspid atresia, and 1 atrioventricular septal defect. The surgical interventions included a right Blalock-Taussig (BT) shunts, 3 left BT shunts, 1 bidirectional Glenn shunt, 4 central aortopulmonary shunts, 1 Potts shunt, Waterston shunt, 8 repairs of tetralogy of Fallot, 7 and Jatene procedures. MRI technique. MRI was performed on a 1.5 T imager (Signa, GE Medical Systems, Milwaukee, Wis.), or a 0.35 T system (MT/T, Diasonics, Milpitas, Calif.). For children small enough to fit, a head coil was used. Spin-echo images were acquired with electrocardiographic gating to every heart beat (repetition time [TR] = RR interval; echo time [TEJ = 20 to 30 msec with respiratory compensation). In all
American
Heart Journal
patients, transaxial images with 5 to 10 mm section thickness were obtained to cover the entire thorax. In some patients in the last 2 years of the study, additional 3 mm thick
axial sections were obtained through the region of the main and central pulmonary arteries. Additional coronal and sagittal images were obtained as needed to visualize postsurgical changes. In several of the patients, ECG-referenced tine gradient echo sequences were also acquired. The MRI studies were performed for clinical indications. The studies were done in 560 minutes to conform to the schedule of a clinical imager. Echocardiography technique. graphic techniques were used with struments as available: ATL Mark 8 with 3 or 5 MHz transducers
Standard echocardiodifferent ultrasound in500,600, and Ultramark (Advanced Technology Laboratory, Bothell, Wash.); HP 77020A and 1000 with 3.5 or 5 MHz transducers (Hewlett-Packard, Andover, Mass.);
and Acuson Computed Sonography 128 x P5 with 3,5, or 7 MHz transducers (Acuson, Mountain View, Calif.). Twodimensional echocardiography, continuous wave and pulsed Doppler, and color Doppler imaging were used. Images were acquired in standard echocardiographic planes with additional views if and when needed to evaluate postsurgical changes. Evaluation of the pulmonary artery and anatomic structures The pulmonary arterial tree and pulmonary outflow tract were subdivided into the following five parts: (1) the outflow tract, which included the right ventricular outflow tract, pulmonary valve, and proximal portion of the main pulmonary artery; (2) the pulmonary artery confluence, which included the distal main pulmonary artery and bifurcation; (3) the right pulmonary artery; (4) the left pulmonary artery; and (5) surgical shunts. The MRIs, echocardiograms, and angiograms were evaluated for the presence or absence of a main pulmonary artery, the presence or absence of a pulmonary artery confluence, the presence of a right and left pulmonary artery, and stenosis or hypoplasia of the main, right, or left pulmonary artery or confluence. The patency and presence of stenoses in shunts were identified if possible. Data analysis. The MRI studies were retrospectively reviewed in all 33 patients by two reviewers (A.D., S.M.). The echocardiograms were retrospectively reviewed in all 33 patients by another independent reviewer (A.B.). The results of the MRI and echocardiogram were confirmed by surgical and angiographic findings. Angiogram reports were available in all 33 patients. In 16 of 33 patients the angiocardiograms were independently reviewed by two observers (A.D., L.W.). The criteria for the comparison of MRI and echocardiography were as follows: MRI was considered equal to echocardiography (M = E) when they demonstrated abnormalities or the anatomy equally well. Echocardiographywas considered superior to MRI (E > M) when the echocardiography demonstrated an abnormality not or incompletely diagnosed by MR. MRI was considered superior to echocardiography (M > E) when MRI demonstrated abnormalities not or incompletely diagnosed by echocardiography. When an echocardiographic study was unable to or failed to reveal part of the pulmonary arterial
Volume 126, Number 6, PM 1 American Heart Journal
Duerinckx et al.
Fig. 1. Repair of tetralogy of Fallot (TOF). A, Transaxial spin-echo image with ECG gating demonstrates artery. B and dilated right ventricular outflow tract (rvot) and stenosis (arrow) of origin of left pulmonary C, Oblique coronal planes through origin of right pulmonary artery (RPA; B) and left pulmonary artery (LPA; C) clearly show origins and full lengths of pulmonary arteries. Also shown are aorta (A), right atrium (RA) , left atrium (LA), and left ventricle (L V) . D, Sagittal plane allows visualization of right ventricle fR V) , right ventricular outflow tract, and main pulmonary artery (m).
1141
1142
Duerinckx
et al.
December American
1994
Heart Journal
Fig. 2. A l&month-old boy with small right ventricle and nearly atretic pulmonary outflow tract underwent patch repair of right ventricular outflow tract at birth. Two transaxial images through pulmonary arteries obtained with ECG-gated spin-echo sequence at level of left pulmonary artery (A, arrow) and at level of right pulmonary artery (B, arrow). Pulmonary arteries are hypoplastic.
tree, the MRI was considered superior to the echocardiogram and vice versa. MRI studies always covered the full extent of the pulmonary arterial tree and areas with surgical shunts. The findings of the pulsed and continuous wave Doppler and color Doppler studies were included when echocardiograms were evaluated. Statistical analysis. The differences in performance of echocardiography and MRI in depicting the pulmonary anatomy were compared as follows: for each portion of the anatomy (RVOT, confluence, right and left pulmonary arteries, and shunts) the number of cases in which all abnormalities and the full anatomy were well visualized was tabulated for both echocardiography and MRI. The chisquared test was used to calculate the significance of the difference in the number of cases partially versus totally
visualized with MRI and echocardiography for each portion of the anatomy.g RESULTS Right ventricular outflow tract, pulmonary valve, and proximal main pulmonary artery. In 15 of the 33 pa-
tients, MRI and echocardiography were equally effective (Fig. 1). In 9 of the 33 patients MRI was superior to echocardiography. In 3 of these 9 patients MRI was superior to echocardiography because the echocardiogram did not include this portion of the anatomy. In 9 of the 33 patients echocardiography was superior to MRI. In many instances echocardiography was superior to MRI in the evaluation of the
volume 126, Number 6, Part 1 Amwlcan Heart Journal
Duerinckx
et al.
1143
Fig. 3. A 21/2-year-old girl underwent arterial switch operation (Jatene procedure) for anatomic correction of complete transposition of great arteries. Transaxial spin-echo images with ECG gating show result of arterial switch. Main pulmonary artery (p) lies anterior to aorta (a), and there is severe narrowing of right (A, arrow) and mild narrowing of left (6, arrow) pulmonary arteries.
outflow tract and the pulmonary valve because of the information supplied by Doppler (two-dimensional pulsed and continuous wave and/or color Doppler). The presence of flow through a very stenotic pulmanic valve could sometimes only be appreciated with color Doppler. The chi-squared test showed no significant difference between the performance of MRI and echocardiography (p > 0.1). Right pulmonary artery. MRI and echocardiography performed equally well in 15 of the 33 patients (Figs. 1 through 3). In 18 of the 33 patients, MRI was superior to echocardiography and demonstrated abnormalities and size of the right pulmonary artery better than echocardiography. There was a significant
difference between the performance of MRI and echocardiography (p < 0.001). Left pulmonary artery. MRI and echocardiography performed equally well in 11 of the 33 patients (Figs. 1 through 3). In 22 of the 33 patients, MRI was superior to echocardiography and demonstrated abnormalities and size of the left pulmonary artery better than echocardiography. There was a significant difference between the performance of MRI and echocardiography (p < 0.001). Confluence of pulmonary arteries. Echocardiography demonstrated the absence or presence of confluence of the pulmonary arteries in 29 of 33 patients but failed to demonstrate it in 4 of the 33 patients in the
1144
Duerinckx
December
et al.
American
1994
Heart Journal
Fig. 4. A 2-year-old girl with Down’s syndrome, pulmonary atresia, and ventricular septal defect underwent palliative surgery with right-sided Blalock-Taussig (BT) shunt. Left, Transaxial spin-echo image with ECG gating demonstrates BT shunt (straight arrow) posterior to superior vena cava (curved arrow). Right, Multiplanar reconstruction technique is used to create double-oblique planes along direction of BT shunt (solid white arrows). Also shown are right subclavian artery (curved black arrow), right pulmonary artery (open arrow), right atrium (a), and aortic arch (A).
older age group. MRI showed the absence or presence of confluence in all 33 patients. The chi-squared test showed a moderately significant difference between the performance of MRI and echocardiography (p < 0.05; x2 = 4.26 with 1 degree of freedom). Surgical shunts. Shunts were present in 16 of 33
patients. Surgical shunts were visualized by MRI and/or echocardiography in each of these 16 patients (Fig. 4). In 10 of the 16 patients echocardiography and MRI performed equally well. In 4 of the 16 patients, MR was superior to echocardiography. In 2 of the 16 patients, echocardiography was superior to MRI. There was no significant difference between the performance of MRI and echocardiography. Overall, MRI was superior to echocardiography in 38% of comparisons (33 patients with 5 portions of the pulmonary anatomy each). However, MRI was superior to echocardiography for the detection of abnormalities of the right and left pulmonary arteries in 61% of cases.
DISCUSSION
There has been considerable interest in imaging techniques to determine pulmonary artery size and the configuration of the pulmonary artery to decide on and plan surgical
intervention
in patients
with
cyanosis and decreased pulmonary blood flow. Specifically, after palliative surgical procedures, pulmonary arterial growth with time is unpredictable. It is important to monitor pulmonary arterial size after surgery to guide future total surgical correction and optimal timing
of such intervention
in patients
with
congenital heart disease. The pulmonary arteries can be evaluated by means of x-ray, angiography, pulmonary angioscopy, echocardiography, and MRI. lo Angiography of the pulmonary arteries often requires a subselective injection and wedged pulmonary vein injections. These techniques have their own limitations and, given the high radiation burden and the need to inject iodinated contrast media, they are poorly suited for the
voluma 126, Number 6, Part 1 American HearI Journal
continuous monitoring of pulmonary arterial growth in infants and small children after initial palliative or corrective surgery. Two-dimensional echocardiography is now the most frequently used noninvasive imaging technique in congenital heart disease,ll* I2 but the results of this technique for assessing pulmonary arterial dimensions have been quite variable. A study by Gutgesell et a1.13 describes the high diagnostic effectiveness of echocardiography in demonstrating both right and left pulmonary arteries in a group of 20 patients with pulmonary atresia or severe pulmonary artery stenosis. Most other studies seem to indicate difficulty using echocardiography to evaluate the left pulmonary arteries. In a study by Huhta et al.,ll of 65 patients (aged 16 months to 15 years) with pulmonary atresia with ventricular septal defect, the right pulmonary artery was seen in 55 (85 % ) of 65 patients, but the left pulmonary artery was only visualized in 16 (25 % ) of 65 patients. In another study by Gomes et al.,3 the left pulmonary artery could only be identified in 15 (58%) of 26 patients with pulmonary artery confluence. There appears to be added value of color flow mapping and Doppler echocardiography in determining pulmonary blood supply in infants with pulmonary atresia and ventricular septal defect.14s l5 MRI has been used to evaluate a number of abnormalities of the pulmonary artery, including congenital obstruction of the right ventricular outflow tract,3 pulmonary atresia,21 7 and postsurgical changes to the pulmonary outflow tract or pulmonary arteries.s In a study of 35 patients with congenital obstruction of the right ventricle, MRI was compared to echocardiography and angiography for the visualization of the pulmonary confluence.3 With regard to the presence of a pulmonary confluence, results of MRI and angiography were in agreement in 31(89% ) of 35 patients, and results of echocardiography and angiography were in agreement in 27 (82%) of 33 patients. MRI and angiography were superior to echocardiography for measurements of the left and right pulmonary artery; MRI and angiography showed good correlation between each other. MRI, x-ray angiography, and echocardiography showed good correlations for measurements of the main pulmonary artery. In a study of 10 patients with pulmonary atresia, MRI appeared to be an effective noninvasive technique for morphologic evaluation of the right ventricular outflow tract, the size and course of the central pulmonary vessels, and the source of the collateral supply to the lunge2 MRI also showed surgical shunts and associated cardiac defects. The current study showed that echocardiography
Lherinc~kx
et al.
1145
was roughly equal to MRI for the evaluation of the right ventricular outflow tract and pulmonary valve abnormalities. MRI was superior to echocardiography in demonstrating the confluence of the central pulmonary arteries (p < 0.05). MRI was also superior to echocardiography in evaluating the right and left pulmonary branches (p < 0.001). MRI was superior in 22 (67 % ) of 33 patients for the evaluation of the left pulmonary artery and in 18 (55 % ) of 33 patients for the evaluation of the right pulmonary artery. MRI and echocardiography performed equally in demonstrating a variety of pulmonary shunts. However, it must be recognized that the better performance of MRI in the present study may have been influenced by the fact that during the course of this study some patients may have been referred for an MRI study when echocardiography failed to provide complete evaluation of the pulmonary circulation. Although not routinely used in the current study because of time constraints in these clinically referred cases, velocity-encoded tine MRI can also provide quantitation of flow, including differential flow in the right and left pulmonary arteries.16v20 Recent reports have demonstrated the validity of using phase contrast MRI to quantify pulmonary arterial flow in normal patients and in patients with pulmonary hypertension17 and lung transplant.ations.‘l The addition of flow quantification to anatomic imaging may considerably augment the performance of MRI in the future. In addition, new developments in ultrafast MRI of pulmonary arteries and threedimensional visualization of the pulmonary arterial tree may have an impact on future usage of MRI for visualization of pulmonary arterial abnormalities.22T23 In conclusion, this study shows that MRI is effective for monitoring the morphologic changes in the pulmonary arterial tree after surgery and may be the preferred noninvasive test, for 1his purpose. REFERENCES
Moulton AL, Malm JR. Pulmonary stenosis, pulmonary atresia, single pulmonary artery, and aneurysm of the pulmonaw artery. In: Baue AE, Geha AS, Hammond CL, Laks H, Naunheim KS, eds. Glenn’s Thoracic and Cardiovascular Surgery. Norwalk, Corm.: Appleton & Lange, 1991. Kersting-Sommerhoff BA, Sechtem IIP, Higgins CB. Evaluation of pulmonary blood supply by nuclear magnetic resonance imaging in pan tients with pulmonary atresia. J Am Co11 Cardiol 1988;11:166-71. Games AS, Lois JF, Williams RG. Pulmonary arteries: MR imaging in patients with congenital obstruction of the right ventricular outflow tract. Radiology 1990,174:51-7. 4. Fletcher BD, Jacobstein MD. MRI of congenital abnormalities of the great arteries. AJR 1986;146:941-8. 5. Formanek AG, Witcofski RL, D’Souza VJ, Link KM, Karstaedt N. MR imaging of the central pulmonary arterial tree in conotruncal malformation. AJR Am J Roentgen01 198X$147:112:-31. 6. Julsrud PR, Ehman RL, Hagler DJ, Illstrup DM. Extracardiac vasculature in candidates for Fontan surgery: MR imaging. Radiology
1146
December
Duerinckx et al.
1989;173:503-6. 7. Rees RS, Sommerville J, Underwood SR, Wright J, Firmin DN, Klipstein RH, Longmore DB. Magnetic resonance imaging of the pulmonary arteries and their systemic connections in pulmonary atresia: cornpar ison with angiographic and surgical findings. Br Heart J 1987;58:621-6. 8. Duerinckx AJ, Rhee JM, Mahboubi S, Higgins SS, Hardy CE, Higgins CB. MR evaluation of the pulmonary arteries after congenital heart surgery: comparison with echocardiography. Circulation 1992;86 (suppl):l-500. 9. Dawson-Saunders B, Trapp RG, eds. Basic and Clinical Biostatistics. Norwalk, Conn.: Appleton & Lange, 1990:148-50. 10. Fedullo PF, Shure D. Pulmonary vascular imaging. Clin Chest Med 1987;8:63-4. 11. Huhta JC, Piehler JM, Tajik AJ, Hagler DJ, Mair DO, Julsrud PR, Seward JP. Two-dimensional echocardiographic detection and measurement of the right pulmonary artery in pulmonary atresia-ventricular septal defect: angiographic and surgical correlation. Am J Cardiol 1982;49:1235-40. 12. Snider AR, Enderlein MA, Teital DF, Juster RJ. Two-dimensional echocardiographic determination of aortic and pulmonary artery sixes from infancy to adulthood in normal subjects. Am J Cardiol1984;53:21824. 13. Gutgesell HP, Huhta JC, Cohen MH, Latson LA. Two-dimensional echocardiographic assessment of pulmonary artery and aortic arch anatomy in cyanotic infants. Am Co11 Cardiol 1984;4:1242-6. 14. Smyllie JH, Sutherland GR, Keeton BR. The value of Doppler color flow mapping in determining pulmonary blood supply in infants with pulmonary atresia with ventricular septal defect. J Am Co11 Cardiol 1989;14:1759-65. 15. Takarad M, Miyaxawa Y, Yasui S, Horigome H. Clinical usefulness of color-coded two-dimensional Doppler echocardiography in congenital heart disease. In: Doyle EF, eds. Pediatric Cardiology. New York:
AVAILABILITY
OF JOURNAL
BACK
1994
Springer Verlag, 1986. 16. Spritzer CE, Pelt NJ, Lee JN, Evans AJ, Sostman HD, Riederer SJ. Rapid MR imaging of blood flow with a phase-sensitive, limited-flipangle, gradient recalled pulse sequence: preliminary experience. Radiology 1990;176:255-62. 17. Kondo C, Caputo GR, Masui T, Foster E, O’Sullivan M, Stulbarg MS, Golden J, Chatterjee K, Higgins CB. Pulmonary hypertension: pulmonary flow quantification and flow profile analysis with velocity-encoded tine MR imaging. Radiology 1992$83:751-S. 18. Caputo GR, Kondo C, Masui T, Garaci SJ, Foster E, O’Sullivan M, Higgins CB. Right and left lung perfusion: in vitro and in viva validation with oblique-angle, velocity-encoded tine MR imaging. Radiology 1991;180:693-8. 19. Martinez JE, Mohiaddin RH, Kilner PJ, Koktee K, Rees RSO, Somerville J, Longmore DB. Obstruction in extracardiac ventriculopulmonary conduits: value of nuclear magnetic resonance imaging with velocity mapping and Doppler echocardiography. J Am Co11 Cardiol1992;20:33844. 20. Kondo C, Caputo GR, Semelka R, Shimakawa A, Higgins CB. Right and left ventricular stroke volume measurements with velocity encoded tine NMR imaging. AJR Am J Roentgen01 1991;157:9-16. 21. Mohiaddin RH, Pax R, Theodoropoulos S, Firmin DN, Longmore DB, Yacoub MH. Magnetic resonance characterization of pulmonary arterial blood flow after single lung transplantation. J Thorac Cardiovasc Surg 1991;101:1016-23. 22. Foo TKF, MacFall JR, Hayes CE, Sostman HD, Slayman BE. Pulmonary vasculature: single breath-hold MR imaging with phased array coils. Radiology 1992;183:476-7. 23. Wielopolski PA, Hqaacke EM, Adler LP. Three-dimensional MR imaging of the pulmonary vasculature: preliminary experience. Radiology 1992;183:465-72.
ISSUES
As a service to our subscribers, copies of back issues of the AMERICAN HEART JOURNAL for the preceding 5 years are maintained and are available for purchase from the publisher, Mosby, at a cost of $10.00 per issue. The following quantity discounts are available: 25% off on quantities of 12 to 23, and one-third off on quantities of 24 or more. Please writs to Mosby-Year Book, Inc., Subscription Services, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, or call (800)453-4361 or (314)453-4351 for information on availability of particular issues. If unavailable from the publisher, photocopies of complete issues are available from University Microfilms International, 300 N. Zeeb Rd., Ann Arbor, MI 48106, (313)761-4796.
Andre J. Duerinckx, MD, PhD, al b Lewis Wexler, MD,” Anirban Banerjee, MD,“, d Sarah S. Higgins, PhDF Christian E. Hardy, MD: Gregg Helton, MD,” John Rhee, MD,” Soroosh Mahboubi, MD,f and Charles B. Higgins, MDa San Francisco, Los Angeles, Stanford, Philadelphia, Pa.
and Oakland,
Palliative and corrective operations for the treatment of cyanotic congenital heart disease frequently involve or potentially influence the size of the pulmonary arteries. After surgery it is usually necessary to sequentially evaluate the pulmonary arteries to detect occult stenoses or to monitor growth of the pulmonary arteries in response to increased blood fl0w.l In patients with pulmonary atresia and venFrom the *Department of Radiology, Magnetic Resonance Imaging Section, University of California, San Francisco; the bRadiology Service, Magnetic Resonance Imaging, Veterans Administration Medical Center, Los Angeles; the CDepartment of Radiology, Stanford University; the dDivision of Cardiology, Children’s Hospital Medical Center, Cincinnati; ePediatric Cardiology, Children’s Hospital Oakland, and the department of Radiology, The Children’s Hospital of Philadelphia. Received
for publication
Jan. 20, 1994; accepted
March
7, 1994.
Reprint requests: Charles B. Higgins, MD, Department of Radiology, MRI Section, University of California, 505 Parnassus Ave., Suite L306, San Francisco, CA 94143-0628. Copyright @ 1994 by Mosby-Year 0002~8703/94/$3.00 + 0 4/l/58600
Book,
Inc.
Calif.,
Cincinnati,
Ohio, and
tricular septal defect, the size of the pulmonary arteries is a prime determinant in the appropriateness of a one-stage surgical correction or the need for preliminary palliative surgical procedures to stimulate pulmonary arterial growth that is compatible with later repair. After initial palliative surgery, evaluation of interval growth of pulmonary arteries is also needed before a method of complete correction is chosen. The techniques currently used to assess the pulmonary arteries are angiocardiography, echocardiography (echo), and magnetic resonance imaging (MRI). Angiography has limitations in visualizing pulmonary arteries in patients with pulmonary atresia and predominate flow to the lungs by aortopulmonary collaterals. Echocardiography is sometimes limited in demonstrating the pulmonary arteries by poor sound transmission because of overlying lung, especially in older children. MRI is another noninvasive method for evaluating congenital heart disease, in1139
1140
December 1994
Duerinckx et al.
eluding palliative systemic pulmonary artery shunts, central and proximal pulmonary arteries, and greatvessel abnormalities.2-8 MRI has been shown to be effective in the evaluation of patients with obstruction of the right ventricular outflow tract3 and in patients with pulmonary atresia.2T 7 Although previous studies2-7 have shown the effectiveness of MRI in evaluating the pulmonary arteries, only a single study3 shows comparisons between MRI and echo for the evaluation of pulmonary arterial abnormalities. No studies have provided comparative data specifically in patients after surgery involving the pulmonary arteries. The purpose of the present study was to compare MRI to echo in the evaluation of the morphologic changes of the pulmonary arteries in patients with congenital heart disease after surgery that might influence the size and/or configuration of the pulmonary arteries. Results of the noninvasive modalities were confirmed from angiography and/or surgical findings. METHODS Study population. Seventy-two patients with congenital heart disease in whom prior surgery had involved the pulmonary arteries were referred for MRI from 1988 to 1992. Of the original 72 patients, 33 had echocardiograms performed within 12 months of the MRI study that were available for review, and no intervening surgery or treatment had occurred between the two studies. The ages of the 33 patients ranged from 1 month to 26 years (mean 4.5 years). Patients were excluded from the study if (1) there was intervening surgery between the dates of the MRI and echocardiography studies; (2) one of the studies was missing or not available for review; or (3) there was a >12-month interval between the two studies. The average time difference between the time of the MRI and echocardiography was 2.5 months (range 0 days to 12 months). Major cardiovascular anomalies in this group of 33 patients were varied: 8 had tetralogy of Fallot, 9 transposition of the great arteries, 2 pulmonary atresia with intact ventricular septum, 3 single ventricle, 2 hypoplastic left heart syndrome, 3 truncus arteriosus, 1 pulmonic stenosis, 2 double-outlet right ventricle, 1 Ebstein’s anomaly, 1 tricuspid atresia, and 1 atrioventricular septal defect. The surgical interventions included a right Blalock-Taussig (BT) shunts, 3 left BT shunts, 1 bidirectional Glenn shunt, 4 central aortopulmonary shunts, 1 Potts shunt, Waterston shunt, 8 repairs of tetralogy of Fallot, 7 and Jatene procedures. MRI technique. MRI was performed on a 1.5 T imager (Signa, GE Medical Systems, Milwaukee, Wis.), or a 0.35 T system (MT/T, Diasonics, Milpitas, Calif.). For children small enough to fit, a head coil was used. Spin-echo images were acquired with electrocardiographic gating to every heart beat (repetition time [TR] = RR interval; echo time [TEJ = 20 to 30 msec with respiratory compensation). In all
American
Heart Journal
patients, transaxial images with 5 to 10 mm section thickness were obtained to cover the entire thorax. In some patients in the last 2 years of the study, additional 3 mm thick
axial sections were obtained through the region of the main and central pulmonary arteries. Additional coronal and sagittal images were obtained as needed to visualize postsurgical changes. In several of the patients, ECG-referenced tine gradient echo sequences were also acquired. The MRI studies were performed for clinical indications. The studies were done in 560 minutes to conform to the schedule of a clinical imager. Echocardiography technique. graphic techniques were used with struments as available: ATL Mark 8 with 3 or 5 MHz transducers
Standard echocardiodifferent ultrasound in500,600, and Ultramark (Advanced Technology Laboratory, Bothell, Wash.); HP 77020A and 1000 with 3.5 or 5 MHz transducers (Hewlett-Packard, Andover, Mass.);
and Acuson Computed Sonography 128 x P5 with 3,5, or 7 MHz transducers (Acuson, Mountain View, Calif.). Twodimensional echocardiography, continuous wave and pulsed Doppler, and color Doppler imaging were used. Images were acquired in standard echocardiographic planes with additional views if and when needed to evaluate postsurgical changes. Evaluation of the pulmonary artery and anatomic structures The pulmonary arterial tree and pulmonary outflow tract were subdivided into the following five parts: (1) the outflow tract, which included the right ventricular outflow tract, pulmonary valve, and proximal portion of the main pulmonary artery; (2) the pulmonary artery confluence, which included the distal main pulmonary artery and bifurcation; (3) the right pulmonary artery; (4) the left pulmonary artery; and (5) surgical shunts. The MRIs, echocardiograms, and angiograms were evaluated for the presence or absence of a main pulmonary artery, the presence or absence of a pulmonary artery confluence, the presence of a right and left pulmonary artery, and stenosis or hypoplasia of the main, right, or left pulmonary artery or confluence. The patency and presence of stenoses in shunts were identified if possible. Data analysis. The MRI studies were retrospectively reviewed in all 33 patients by two reviewers (A.D., S.M.). The echocardiograms were retrospectively reviewed in all 33 patients by another independent reviewer (A.B.). The results of the MRI and echocardiogram were confirmed by surgical and angiographic findings. Angiogram reports were available in all 33 patients. In 16 of 33 patients the angiocardiograms were independently reviewed by two observers (A.D., L.W.). The criteria for the comparison of MRI and echocardiography were as follows: MRI was considered equal to echocardiography (M = E) when they demonstrated abnormalities or the anatomy equally well. Echocardiographywas considered superior to MRI (E > M) when the echocardiography demonstrated an abnormality not or incompletely diagnosed by MR. MRI was considered superior to echocardiography (M > E) when MRI demonstrated abnormalities not or incompletely diagnosed by echocardiography. When an echocardiographic study was unable to or failed to reveal part of the pulmonary arterial
Volume 126, Number 6, PM 1 American Heart Journal
Duerinckx et al.
Fig. 1. Repair of tetralogy of Fallot (TOF). A, Transaxial spin-echo image with ECG gating demonstrates artery. B and dilated right ventricular outflow tract (rvot) and stenosis (arrow) of origin of left pulmonary C, Oblique coronal planes through origin of right pulmonary artery (RPA; B) and left pulmonary artery (LPA; C) clearly show origins and full lengths of pulmonary arteries. Also shown are aorta (A), right atrium (RA) , left atrium (LA), and left ventricle (L V) . D, Sagittal plane allows visualization of right ventricle fR V) , right ventricular outflow tract, and main pulmonary artery (m).
1141
1142
Duerinckx
et al.
December American
1994
Heart Journal
Fig. 2. A l&month-old boy with small right ventricle and nearly atretic pulmonary outflow tract underwent patch repair of right ventricular outflow tract at birth. Two transaxial images through pulmonary arteries obtained with ECG-gated spin-echo sequence at level of left pulmonary artery (A, arrow) and at level of right pulmonary artery (B, arrow). Pulmonary arteries are hypoplastic.
tree, the MRI was considered superior to the echocardiogram and vice versa. MRI studies always covered the full extent of the pulmonary arterial tree and areas with surgical shunts. The findings of the pulsed and continuous wave Doppler and color Doppler studies were included when echocardiograms were evaluated. Statistical analysis. The differences in performance of echocardiography and MRI in depicting the pulmonary anatomy were compared as follows: for each portion of the anatomy (RVOT, confluence, right and left pulmonary arteries, and shunts) the number of cases in which all abnormalities and the full anatomy were well visualized was tabulated for both echocardiography and MRI. The chisquared test was used to calculate the significance of the difference in the number of cases partially versus totally
visualized with MRI and echocardiography for each portion of the anatomy.g RESULTS Right ventricular outflow tract, pulmonary valve, and proximal main pulmonary artery. In 15 of the 33 pa-
tients, MRI and echocardiography were equally effective (Fig. 1). In 9 of the 33 patients MRI was superior to echocardiography. In 3 of these 9 patients MRI was superior to echocardiography because the echocardiogram did not include this portion of the anatomy. In 9 of the 33 patients echocardiography was superior to MRI. In many instances echocardiography was superior to MRI in the evaluation of the
volume 126, Number 6, Part 1 Amwlcan Heart Journal
Duerinckx
et al.
1143
Fig. 3. A 21/2-year-old girl underwent arterial switch operation (Jatene procedure) for anatomic correction of complete transposition of great arteries. Transaxial spin-echo images with ECG gating show result of arterial switch. Main pulmonary artery (p) lies anterior to aorta (a), and there is severe narrowing of right (A, arrow) and mild narrowing of left (6, arrow) pulmonary arteries.
outflow tract and the pulmonary valve because of the information supplied by Doppler (two-dimensional pulsed and continuous wave and/or color Doppler). The presence of flow through a very stenotic pulmanic valve could sometimes only be appreciated with color Doppler. The chi-squared test showed no significant difference between the performance of MRI and echocardiography (p > 0.1). Right pulmonary artery. MRI and echocardiography performed equally well in 15 of the 33 patients (Figs. 1 through 3). In 18 of the 33 patients, MRI was superior to echocardiography and demonstrated abnormalities and size of the right pulmonary artery better than echocardiography. There was a significant
difference between the performance of MRI and echocardiography (p < 0.001). Left pulmonary artery. MRI and echocardiography performed equally well in 11 of the 33 patients (Figs. 1 through 3). In 22 of the 33 patients, MRI was superior to echocardiography and demonstrated abnormalities and size of the left pulmonary artery better than echocardiography. There was a significant difference between the performance of MRI and echocardiography (p < 0.001). Confluence of pulmonary arteries. Echocardiography demonstrated the absence or presence of confluence of the pulmonary arteries in 29 of 33 patients but failed to demonstrate it in 4 of the 33 patients in the
1144
Duerinckx
December
et al.
American
1994
Heart Journal
Fig. 4. A 2-year-old girl with Down’s syndrome, pulmonary atresia, and ventricular septal defect underwent palliative surgery with right-sided Blalock-Taussig (BT) shunt. Left, Transaxial spin-echo image with ECG gating demonstrates BT shunt (straight arrow) posterior to superior vena cava (curved arrow). Right, Multiplanar reconstruction technique is used to create double-oblique planes along direction of BT shunt (solid white arrows). Also shown are right subclavian artery (curved black arrow), right pulmonary artery (open arrow), right atrium (a), and aortic arch (A).
older age group. MRI showed the absence or presence of confluence in all 33 patients. The chi-squared test showed a moderately significant difference between the performance of MRI and echocardiography (p < 0.05; x2 = 4.26 with 1 degree of freedom). Surgical shunts. Shunts were present in 16 of 33
patients. Surgical shunts were visualized by MRI and/or echocardiography in each of these 16 patients (Fig. 4). In 10 of the 16 patients echocardiography and MRI performed equally well. In 4 of the 16 patients, MR was superior to echocardiography. In 2 of the 16 patients, echocardiography was superior to MRI. There was no significant difference between the performance of MRI and echocardiography. Overall, MRI was superior to echocardiography in 38% of comparisons (33 patients with 5 portions of the pulmonary anatomy each). However, MRI was superior to echocardiography for the detection of abnormalities of the right and left pulmonary arteries in 61% of cases.
DISCUSSION
There has been considerable interest in imaging techniques to determine pulmonary artery size and the configuration of the pulmonary artery to decide on and plan surgical
intervention
in patients
with
cyanosis and decreased pulmonary blood flow. Specifically, after palliative surgical procedures, pulmonary arterial growth with time is unpredictable. It is important to monitor pulmonary arterial size after surgery to guide future total surgical correction and optimal timing
of such intervention
in patients
with
congenital heart disease. The pulmonary arteries can be evaluated by means of x-ray, angiography, pulmonary angioscopy, echocardiography, and MRI. lo Angiography of the pulmonary arteries often requires a subselective injection and wedged pulmonary vein injections. These techniques have their own limitations and, given the high radiation burden and the need to inject iodinated contrast media, they are poorly suited for the
voluma 126, Number 6, Part 1 American HearI Journal
continuous monitoring of pulmonary arterial growth in infants and small children after initial palliative or corrective surgery. Two-dimensional echocardiography is now the most frequently used noninvasive imaging technique in congenital heart disease,ll* I2 but the results of this technique for assessing pulmonary arterial dimensions have been quite variable. A study by Gutgesell et a1.13 describes the high diagnostic effectiveness of echocardiography in demonstrating both right and left pulmonary arteries in a group of 20 patients with pulmonary atresia or severe pulmonary artery stenosis. Most other studies seem to indicate difficulty using echocardiography to evaluate the left pulmonary arteries. In a study by Huhta et al.,ll of 65 patients (aged 16 months to 15 years) with pulmonary atresia with ventricular septal defect, the right pulmonary artery was seen in 55 (85 % ) of 65 patients, but the left pulmonary artery was only visualized in 16 (25 % ) of 65 patients. In another study by Gomes et al.,3 the left pulmonary artery could only be identified in 15 (58%) of 26 patients with pulmonary artery confluence. There appears to be added value of color flow mapping and Doppler echocardiography in determining pulmonary blood supply in infants with pulmonary atresia and ventricular septal defect.14s l5 MRI has been used to evaluate a number of abnormalities of the pulmonary artery, including congenital obstruction of the right ventricular outflow tract,3 pulmonary atresia,21 7 and postsurgical changes to the pulmonary outflow tract or pulmonary arteries.s In a study of 35 patients with congenital obstruction of the right ventricle, MRI was compared to echocardiography and angiography for the visualization of the pulmonary confluence.3 With regard to the presence of a pulmonary confluence, results of MRI and angiography were in agreement in 31(89% ) of 35 patients, and results of echocardiography and angiography were in agreement in 27 (82%) of 33 patients. MRI and angiography were superior to echocardiography for measurements of the left and right pulmonary artery; MRI and angiography showed good correlation between each other. MRI, x-ray angiography, and echocardiography showed good correlations for measurements of the main pulmonary artery. In a study of 10 patients with pulmonary atresia, MRI appeared to be an effective noninvasive technique for morphologic evaluation of the right ventricular outflow tract, the size and course of the central pulmonary vessels, and the source of the collateral supply to the lunge2 MRI also showed surgical shunts and associated cardiac defects. The current study showed that echocardiography
Lherinc~kx
et al.
1145
was roughly equal to MRI for the evaluation of the right ventricular outflow tract and pulmonary valve abnormalities. MRI was superior to echocardiography in demonstrating the confluence of the central pulmonary arteries (p < 0.05). MRI was also superior to echocardiography in evaluating the right and left pulmonary branches (p < 0.001). MRI was superior in 22 (67 % ) of 33 patients for the evaluation of the left pulmonary artery and in 18 (55 % ) of 33 patients for the evaluation of the right pulmonary artery. MRI and echocardiography performed equally in demonstrating a variety of pulmonary shunts. However, it must be recognized that the better performance of MRI in the present study may have been influenced by the fact that during the course of this study some patients may have been referred for an MRI study when echocardiography failed to provide complete evaluation of the pulmonary circulation. Although not routinely used in the current study because of time constraints in these clinically referred cases, velocity-encoded tine MRI can also provide quantitation of flow, including differential flow in the right and left pulmonary arteries.16v20 Recent reports have demonstrated the validity of using phase contrast MRI to quantify pulmonary arterial flow in normal patients and in patients with pulmonary hypertension17 and lung transplant.ations.‘l The addition of flow quantification to anatomic imaging may considerably augment the performance of MRI in the future. In addition, new developments in ultrafast MRI of pulmonary arteries and threedimensional visualization of the pulmonary arterial tree may have an impact on future usage of MRI for visualization of pulmonary arterial abnormalities.22T23 In conclusion, this study shows that MRI is effective for monitoring the morphologic changes in the pulmonary arterial tree after surgery and may be the preferred noninvasive test, for 1his purpose. REFERENCES
Moulton AL, Malm JR. Pulmonary stenosis, pulmonary atresia, single pulmonary artery, and aneurysm of the pulmonaw artery. In: Baue AE, Geha AS, Hammond CL, Laks H, Naunheim KS, eds. Glenn’s Thoracic and Cardiovascular Surgery. Norwalk, Corm.: Appleton & Lange, 1991. Kersting-Sommerhoff BA, Sechtem IIP, Higgins CB. Evaluation of pulmonary blood supply by nuclear magnetic resonance imaging in pan tients with pulmonary atresia. J Am Co11 Cardiol 1988;11:166-71. Games AS, Lois JF, Williams RG. Pulmonary arteries: MR imaging in patients with congenital obstruction of the right ventricular outflow tract. Radiology 1990,174:51-7. 4. Fletcher BD, Jacobstein MD. MRI of congenital abnormalities of the great arteries. AJR 1986;146:941-8. 5. Formanek AG, Witcofski RL, D’Souza VJ, Link KM, Karstaedt N. MR imaging of the central pulmonary arterial tree in conotruncal malformation. AJR Am J Roentgen01 198X$147:112:-31. 6. Julsrud PR, Ehman RL, Hagler DJ, Illstrup DM. Extracardiac vasculature in candidates for Fontan surgery: MR imaging. Radiology
1146
December
Duerinckx et al.
1989;173:503-6. 7. Rees RS, Sommerville J, Underwood SR, Wright J, Firmin DN, Klipstein RH, Longmore DB. Magnetic resonance imaging of the pulmonary arteries and their systemic connections in pulmonary atresia: cornpar ison with angiographic and surgical findings. Br Heart J 1987;58:621-6. 8. Duerinckx AJ, Rhee JM, Mahboubi S, Higgins SS, Hardy CE, Higgins CB. MR evaluation of the pulmonary arteries after congenital heart surgery: comparison with echocardiography. Circulation 1992;86 (suppl):l-500. 9. Dawson-Saunders B, Trapp RG, eds. Basic and Clinical Biostatistics. Norwalk, Conn.: Appleton & Lange, 1990:148-50. 10. Fedullo PF, Shure D. Pulmonary vascular imaging. Clin Chest Med 1987;8:63-4. 11. Huhta JC, Piehler JM, Tajik AJ, Hagler DJ, Mair DO, Julsrud PR, Seward JP. Two-dimensional echocardiographic detection and measurement of the right pulmonary artery in pulmonary atresia-ventricular septal defect: angiographic and surgical correlation. Am J Cardiol 1982;49:1235-40. 12. Snider AR, Enderlein MA, Teital DF, Juster RJ. Two-dimensional echocardiographic determination of aortic and pulmonary artery sixes from infancy to adulthood in normal subjects. Am J Cardiol1984;53:21824. 13. Gutgesell HP, Huhta JC, Cohen MH, Latson LA. Two-dimensional echocardiographic assessment of pulmonary artery and aortic arch anatomy in cyanotic infants. Am Co11 Cardiol 1984;4:1242-6. 14. Smyllie JH, Sutherland GR, Keeton BR. The value of Doppler color flow mapping in determining pulmonary blood supply in infants with pulmonary atresia with ventricular septal defect. J Am Co11 Cardiol 1989;14:1759-65. 15. Takarad M, Miyaxawa Y, Yasui S, Horigome H. Clinical usefulness of color-coded two-dimensional Doppler echocardiography in congenital heart disease. In: Doyle EF, eds. Pediatric Cardiology. New York:
AVAILABILITY
OF JOURNAL
BACK
1994
Springer Verlag, 1986. 16. Spritzer CE, Pelt NJ, Lee JN, Evans AJ, Sostman HD, Riederer SJ. Rapid MR imaging of blood flow with a phase-sensitive, limited-flipangle, gradient recalled pulse sequence: preliminary experience. Radiology 1990;176:255-62. 17. Kondo C, Caputo GR, Masui T, Foster E, O’Sullivan M, Stulbarg MS, Golden J, Chatterjee K, Higgins CB. Pulmonary hypertension: pulmonary flow quantification and flow profile analysis with velocity-encoded tine MR imaging. Radiology 1992$83:751-S. 18. Caputo GR, Kondo C, Masui T, Garaci SJ, Foster E, O’Sullivan M, Higgins CB. Right and left lung perfusion: in vitro and in viva validation with oblique-angle, velocity-encoded tine MR imaging. Radiology 1991;180:693-8. 19. Martinez JE, Mohiaddin RH, Kilner PJ, Koktee K, Rees RSO, Somerville J, Longmore DB. Obstruction in extracardiac ventriculopulmonary conduits: value of nuclear magnetic resonance imaging with velocity mapping and Doppler echocardiography. J Am Co11 Cardiol1992;20:33844. 20. Kondo C, Caputo GR, Semelka R, Shimakawa A, Higgins CB. Right and left ventricular stroke volume measurements with velocity encoded tine NMR imaging. AJR Am J Roentgen01 1991;157:9-16. 21. Mohiaddin RH, Pax R, Theodoropoulos S, Firmin DN, Longmore DB, Yacoub MH. Magnetic resonance characterization of pulmonary arterial blood flow after single lung transplantation. J Thorac Cardiovasc Surg 1991;101:1016-23. 22. Foo TKF, MacFall JR, Hayes CE, Sostman HD, Slayman BE. Pulmonary vasculature: single breath-hold MR imaging with phased array coils. Radiology 1992;183:476-7. 23. Wielopolski PA, Hqaacke EM, Adler LP. Three-dimensional MR imaging of the pulmonary vasculature: preliminary experience. Radiology 1992;183:465-72.
ISSUES
As a service to our subscribers, copies of back issues of the AMERICAN HEART JOURNAL for the preceding 5 years are maintained and are available for purchase from the publisher, Mosby, at a cost of $10.00 per issue. The following quantity discounts are available: 25% off on quantities of 12 to 23, and one-third off on quantities of 24 or more. Please writs to Mosby-Year Book, Inc., Subscription Services, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, or call (800)453-4361 or (314)453-4351 for information on availability of particular issues. If unavailable from the publisher, photocopies of complete issues are available from University Microfilms International, 300 N. Zeeb Rd., Ann Arbor, MI 48106, (313)761-4796.