- Email: [email protected]
Usefulness of Magnetic Resonance Angiography in the Evaluation of Complex Congenital Heart Disease in Newborns and Infants
Usefulness of Magnetic Resonance Angiography in the Evaluation of Complex Congenital Heart Disease in Newborns and Infants Ashwin Prakash, MDa,b,*, Alejandro J. Torres, MDa, Beth F. Printz, MD, PhDa,b, Martin R. Prince, MD, PhDb, and James C. Nielsen, MDc This study evaluated the quality of the visualization of extracardiac thoracic vessels by magnetic resonance angiography (MRA) in young infants with congenital heart disease. Echocardiography is often sufficient in evaluating CHD in young infants. Cardiac catheterization is needed in some instances to evaluate extracardiac thoracic vessels. Extracardiac thoracic vessels can be accurately evaluated using MRA in adults and older children, but image quality in small infants may be limited. Twenty-nine magnetic resonance angiographic scans were performed at a single institution on 28 infants aged <3 months (median 6 days, range 1 to 90 days) with complex CHD in whom imaging was inconclusive by echocardiography. A blinded observer at a different institution graded (from 0 to 3) the quality of the visualization of the main, branch, lobar, and second-generation pulmonary arteries; lobar pulmonary veins; aortopulmonary collaterals; vena cavae; thoracic aorta and its branches; patent ductus arteriosus; and visceral sidedness. The results of MRA were compared with those of x-ray angiography and surgical inspection, when available. The mean image quality grade was >2 for all structures except the second-generation pulmonary arterial branches, for which it was 2. The median total scan duration was 9 minutes (range 3 to 46). Findings were concordant with surgical inspection (n ⴝ 25) and cardiac catheterization (n ⴝ 8) in all subjects. There were no complications. In conclusion, MRA is excellent for the visualization of extracardiac thoracic vessels in young infants with CHD and can be used as an alternative to cardiac catheterization when echocardiography is inconclusive. © 2007 Elsevier Inc. All rights reserved. (Am J Cardiol 2007;100:715–721)
Three-dimensional contrast-enhanced magnetic resonance angiography (MRA) is a noninvasive technique that has been shown to be accurate in the delineation of vascular anatomy in adults and older children and may be used instead of x-ray angiography.1–5 However, the routine use of this technique in young infants (aged ⬍3 months) has been limited by technical problems, including low image signalto-noise ratio, artifacts related to respiratory motion, hypothermia, bolus timing, and the lack of adequately sized receiver coils. The effects of these technical limitations on the quality of the visualization of individual thoracic vessels and on the diagnostic accuracy of MRA in this age group are not known. To define the role of MRA in the evaluation of these infants, it is important to systematically assess the quality of the visualization and diagnostic accuracy of MRA related to each thoracic vascular structure of interest in this patient population. We evaluated the image quality of the pulmonary arterial bed and other extracardiac thoracic vessels by MRA in newborns and young infants with congenital heart disease (CHD).
a
Division of Pediatric Cardiology and bDepartment of Radiology, Columbia University College of Physicians and Surgeons; and cDivision of Pediatric Cardiology, Mount Sinai School of Medicine, New York, New York. Manuscript received January 5, 2007; revised manuscript received and accepted March 7, 2007. *Corresponding author: Tel: 212-305-8261; fax: 212-305-4429. E-mail address: [email protected] (A. Prakash). 0002-9149/07/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2007.03.090
Methods Study population: All infants aged ⬍3 months with CHD who underwent contrast-enhanced MRA at Morgan Stanley Children’s Hospital of New York Presbyterian from March 2004 to February 2005 were included in this retrospective study. The primary diagnosis for each subject is listed in Table 1. The Columbia University Institutional Review Board gave permission for the review of existing data. The decision to perform MRA and/or cardiac catheterization was made by the patient’s cardiologist on a caseby-case basis. Magnetic resonance angiographic protocol: Magnetic resonance angiographic studies were performed using a 1.5-T whole-body scanner (Signa EXCITE; General Electric, Milwaukee, Wisconsin), using a commercially available 4-element head-imaging coil. All infants were sedated, paralyzed, and mechanically ventilated during the examinations. MRA was performed in the coronal orientation using the following imaging parameters: field of view 200 mm, matrix 192 (phase) and 256 (frequency), slice thickness 2 mm, echo time 1.1 ms, repetition time 3.4 to 3.7 ms, flip angle 40°, receiver band width 31.3 kHz, number of signal averages 1, sequential phase ordering, number of dynamics 3, and acquisition time 25 to 35 s/dynamic. Gadopentate dimeglumine (Magnevist; Berlex Laboratories, Seattle, Washington) (0.4 to 0.5 mmol/kg) followed by a normal saline flush (1 cm3/kg) was manually injected in a peripheral or central vein in the upper or lower extremity at a rate of approximately 0.5 to 1 cm3/s. After the suspension of www.AJConline.org
716
The American Journal of Cardiology (www.AJConline.org)
Table 1 Primary cardiac diagnoses
were graded separately. The aortopulmonary collaterals were graded as a group in each subject.
Diagnosis
n
Single ventricle physiology Pulmonary venous anomaly Scimitar syndrome Tetralogy of Fallot, pulmonary atresia Left pulmonary artery sling Situs inversus, coarctation of aorta Truncus arteriosus with isolated left pulmonary artery Tetralogy of Fallot, absent pulmonary valve
10 4 4 5 2 1 1 1
Table 2 Image quality grading scale Grade 0 1 2 3 A
Description Structure not seen due to poor image quality Structure seen with indistinct margins, image quality insufficient for diagnosis Structure seen with indistinct margins, image quality sufficient for diagnosis but not for measurement Structure seen with distinct margins, image quality sufficient for diagnosis and measurement Structure absent
mechanical ventilation, imaging was initiated approximately 5 to 10 seconds after the start of the injection, taking into consideration the site of the intravenous line and the length of tubing. Three sets of images (dynamics) were obtained, with ventilator breaths delivered for approximately 5 seconds between each period of apnea. The number of dynamics obtained during each period of apnea was based on the infant’s ability to tolerate apnea and varied from 1 to 2. Oxygen saturation, blood pressure, respiratory rate, and heart rate were monitored noninvasively throughout the procedure. Magnetic resonance angiographic image analysis for image quality: Magnetic resonance angiographic images were electronically deidentified and transferred to a blinded observer (JCN) who was an expert in the magnetic resonance imaging of CHD at a different institution (Mount Sinai Medical Center). This blinded observer performed all image analyses using a commercially available computer workstation and software (Advantage Windows 4.0; General Electric). Cardiac and extracardiac vascular anatomy was defined and graded by constructing user-defined subvolume maximal-intensity projections and multiplanar reformatted images from the 3-dimensional data sets. The quality of the visualization of the pulmonary arterial bed, pulmonary and systemic veins, thoracic aorta and its branches, aortopulmonary collaterals, patent ductus arteriosus, and the sidedness of the abdominal viscera was assessed semiquantitatively using the grading scale listed in Table 2. The pulmonary artery branch to each lung and further branches to each lobe of the lung were graded separately, and the second-generation branches for each lung were graded as a group. The pulmonary vein branches draining each lung lobe were graded separately. The subclavian, carotid, and internal mammary arteries on each side
Comparison with findings of x-ray angiography and surgical inspection: The specific diagnostic questions at referral for each magnetic resonance angiographic scan were determined from the official magnetic resonance angiographic report. Subjects in whom specific confirmation of the diagnostic question by x-ray angiography or surgical inspection (performed within 3 months of MRA) was available were identified. For these subjects, the magnetic resonance angiographic diagnoses for each anatomic segment in question were recorded by an observer who is an expert in the magnetic resonance imaging of CHD (AP). For all subjects, the magnetic resonance angiographic diagnosis of the anatomic segment in question preceded the catheterization or surgical procedure. X-ray angiographic images for these subjects were reviewed by an expert in congenital cardiac catheterization (AJT), and all vascular diagnoses were recorded. The findings of surgical inspection were obtained from the official operative reports. The accuracy of magnetic resonance angiographic diagnosis related to the diagnostic question at referral was thus determined. In addition, official magnetic resonance angiographic, x-ray angiographic, and surgical reports were carefully searched for any other diagnostic discrepancies, which were recorded. Results Subjects and diagnostic questions: Twenty-nine contrast-enhanced magnetic resonance angiographic scans were performed on 28 infants aged ⬍3 months (median 6 days, range 1 to 90 days; median weight 3 kg, range 2.1 to 3.9) during the 1-year study period. The results of MRA in the 4 patients with scimitar syndrome have been reported previously.6 All subjects had undergone echocardiography before MRA. The diagnostic questions at referral for MRA are listed in Table 3. MRA and sedation: All subjects tolerated the suspension of respiration for a median of 25 seconds (range 20 to 48) without significant desaturation or change in heart rate. The median total duration of scanning was 9 minutes (range 3 to 46). There were no complications. Magnetic resonance angiographic image quality: Image quality grades for analyzed structures are listed in Table 4. Overall magnetic resonance angiographic image quality was excellent in every subject. The mean image quality grade was ⬎2 for all structures except the second-generation pulmonary arterial branches, for which it was 2. The image quality grade was ⱖ2 (adequate for diagnosis) for every structure in every subject except for the secondgeneration pulmonary arterial branches. The image quality grade was ⬍2 (inadequate for diagnosis) for the secondgeneration pulmonary arterial branches in only 1 subject with tetralogy of Fallot with pulmonary atresia and severely hypoplastic native pulmonary arteries. Examples of maximum intensity projections reconstructed from magnetic resonance angiographic data sets are shown in Figures 1 to 4. Comparison with findings of x-ray angiography and surgical inspection: The results of x-ray angiography (n ⫽ 8) and/or surgical inspection (n ⫽ 25) were available for 23
Congenital Heart Disease/MRA of Heart Disease in Infancy
717
Table 3 Accuracy of magnetic resonance angiographic assessment of referral diagnostic questions Anatomic Segment for Which MRA Evaluation Was Requested
No. of MRA Scans
Diagnostic Confirmation of Anatomic Segment by Angiography/Surgery Available (n)
No. of Scans in Which MRA Diagnosis Was Concordant With Angiography/Surgery (n)
20 12 13 7 2 3 1
16 11 9 6 2 2 1
16 11 9 6 2 2 1
Pulmonary artery anatomy Pulmonary venous anatomy Aortopulmonary collaterals Anatomy of aortic arch, its branches, or ductus arteriosus Surgical psuedoaneurysm Systemic venous anatomy Airway compression
Table 4 Image quality grades Structure Analyzed
PAs Main PA Proximal PA branches Lobar PA branches Second-generation PA branches PVs Lobar pulmonary veins Anomalous PV confluence Anomalous vertical PV Aortopulmonary collateral arteries Superior vena cavae Right sided Left sided Inferior vena cava Aorta and branches Ascending aorta Transverse aortic arch Isthmus Thoracic descending aorta Carotid and subclavian arteries Internal mammary arteries Patent ductus arteriosus Visceral sidedness Stomach Liver Spleen
Image Quality Grade
No. of Scans in Which Structure Was Mean ⫾ SD Range Present 3⫾0 3⫾0 2.8 ⫾ 0.6 2 ⫾ 0.6
3–3 3–3 2–3 0–3
26 29 29 29
2.9 ⫾ 0.1 3⫾0 3⫾0 2.7 ⫾ 0.5
2–3 3–3 3–3 2–3
29 3 4 6
3⫾0 3⫾0 3⫾0
2–3 2–3 3–3
26 10 28
3⫾0 3⫾0 3⫾0 3⫾0 3⫾0 2.9 ⫾ 0.3 2.9 ⫾ 0.3
3–3 3–3 3–3 3–3 3–3 2–3 2–3
29 27 27 29 29 29 13
3⫾0 3⫾0 3⫾0
3–3 3–3 3–3
29 29 24
PA ⫽ pulmonary artery; PV ⫽ pulmonary vein.
and 28 subjects, respectively. Comparisons of the diagnostic findings of MRA with those of x-ray angiography and surgical inspection, as related to the diagnostic questions at referral, are listed in Table 3. The diagnostic questions at referral were accurately answered by MRA in each subject. No discrepancies were noted between the official magnetic resonance angiographic, x-ray angiographic, and operative reports. For the aortopulmonary collaterals, the number and sites of origin of all collaterals were accurately assessed by MRA. The distribution of branches from the collaterals to individual lung lobes was accurately assessed by MRA in 7 of 9 subjects; in 2 of 9 subjects, there were minor discrepancies between MRA and x-ray angiography in the delineation of the lobar lung distribution of the aortopulmonary
collaterals. The presence of “dual” blood supply to individual lung segments via collaterals and native pulmonary arteries was not assessed by MRA. Indications for cardiac catheterization: Although the diagnostic questions at referral were accurately answered by MRA, in 6 of 28 subjects, additional cardiac catheterization was performed before surgery. The purpose of catheterization was diagnostic in 3 subjects (to confirm the anatomy and lung-segment distribution of the aortopulmonary collaterals in 2 subjects with tetralogy of Fallot with pulmonary atresia and to measure pulmonary artery pressure in 1 subject with scimitar syndrome, ventricular septal defect, and left pulmonary artery sling). In the other 3 subjects, catheterization was performed for interventional procedures (stenting of pulmonary vein stenosis in 1 subject, decompression of the left atrium in 1 subject with hypoplastic left heart syndrome, and embolization of an aortopulmonary collateral in 1 subject with scimitar syndrome). Adverse effects: No immediate adverse effects were noted after the injection of contrast medium. Clinical and laboratory data for a period of ⱖ72 hours were available after 25 of 29 scans in subjects who remained admitted to the hospital. None of these subjects demonstrated an increase in serum creatinine, abnormal hepatic function, or other adverse effects during this period. Three of 29 subjects were discharged home soon after the magnetic resonance imaging scans, and although laboratory data were unavailable, none had a clinical adverse event on follow-up. In 1 subject, the magnetic resonance imaging scan was performed emergently before cardiac surgery. This patient died in the operating room. Discussion The present study demonstrates the utility of MRA in young infants with complex CHD in whom echocardiographic evaluation is incomplete or inconclusive. Because echocardiography provides excellent visualization of intracardiac anomalies, the referral question was frequently related to the extracardiac thoracic vasculature. MRA is especially suited for the evaluation of these anomalies, especially when they occur in combination with each other. After a single intravenous injection of contrast medium, a 3-dimensional data set of the entire vasculature can be obtained rapidly. The present report is the first to systematically study the quality of the visualization of individual extracardiac vascular structures in a group of infants aged ⬍3 months with complex
718
The American Journal of Cardiology (www.AJConline.org)
Figure 1. Maximum-intensity projections (MIPs) from a newborn infant with heterotaxy (asplenia) syndrome, functionally single ventricle, double-outlet right ventricle, pulmonary stenosis, and total anomalous pulmonary venous return showing (A) right superior vena cava (RSVC) and left superior vena cava (LSVC) with a small bridging vein; (B) levocardia with a midline transverse liver; and (C) confluence of right upper (RU), right lower (RL), left upper (LU), and left lower (LL) pulmonary veins behind the left atrium draining superiorly via a vertical vein (VV). A sagittal MIP (D) shows drainage of the VV into the LSVC.
CHD. Image quality was excellent for the diagnosis and measurement of anomalies involving the main, proximal branch, and lobar pulmonary arteries; the systemic and pulmonary veins; the aorta and its branches; and the aortopulmonary collateral arteries and for the determination of abdominal visceral situs. Image quality grade for the second-generation pulmonary arteries was lower on average, and they were not well visualized in 1 subject with tetralogy of Fallot, pulmonary atresia, and hypoplastic native pulmonary arteries. The acquisition parameters used in this series were chosen to provide adequate spatial resolution while maintaining acceptable signal-to-noise characteristics on the basis of our institutional experience, the scanning platform used, and the work of others.7 The field of view and imaging matrix were adjusted to maximize resolution while balancing signal-tonoise characteristics. Parallel-processing techniques such as sensitivity encoding were purposefully not used to maximize the signal-to-noise ratio and because relatively long
periods of apnea are well tolerated by intubated infants. The contrast was injected manually with the “best-estimate” method used for the timing of imaging. Although real-time imaging sequences are now commercially available to allow the timing of contrast bolus, we prefer the best-estimate method in small infants because of their rapid heart rates and relatively fast circulation times. The manual injection of contrast has the advantage of allowing direct monitoring of the subject and communication with the respiratory therapist during the suspension of respiration. In our experience, the site of injection (upper or lower extremity) and the type of venous line (central or peripheral) do not affect the quality of imaging significantly, although this was not evaluated critically in this study. We also believe that a rapid injection rate (0.5 to 1 cm3/s) is important to allow scanning during the “first pass” of the contrast medium because of the fast circulation time in small infants, especially in the presence of intracardiac shunts. The elimination of respiratory
Congenital Heart Disease/MRA of Heart Disease in Infancy
719
Figure 2. Maximum-intensity projections (MIPs) from a newborn infant with hypoplastic left-sided cardiac syndrome after stage 1 palliation with a right ventricle–to–pulmonary artery conduit. Axial (A), sagittal (B), and coronal (C) MIPs show a large psuedoaneurysm (arrow) anterior to the right ventricle–to–pulmonary artery conduit (arrowhead). The reconstructed aortic arch and the aortopulmonary anastomosis are also well seen in a sagittal plane (D). LPA ⫽ left pulmonary artery; RPA ⫽ right pulmonary artery.
motion is important in this group of patients because even slight blurring can make the visualization of small structures difficult in these small-sized subjects. However, this requires endotracheal intubation and paralysis with close monitoring by either an anesthesiologist or an intensivist. These resources may not be available at all institutions, and hence the decision to use intubation and paralysis should be individualized. The median total scanning duration was 9 minutes, showing that in most infants, the referral question can be answered rapidly using MRA. Additional functional assessment, including the quantification of ventricular function and the measurement of blood flow, can be performed using magnetic resonance imaging. However, in most subjects in this series, these data were provided by echocardiography alone or were not thought to be clinically relevant. The optimal dose for MRA of complex CHD in infants aged ⬍3 months is unknown. Dose-finding studies are very difficult to perform in this patient population because of ethical and logistic reasons. This is true for most drugs used
in infants. Accurate diagnoses of the often complex cardiac and vascular anomalies in this patient group require imaging quality of an order that is often limited by the inherently low signal-to-noise ratio related to their small body size. To an extent, the signal-to-noise ratio can be improved with the use of a higher dose of contrast. In this reported cohort, failure to obtain diagnostic images on MRA would have necessitated cardiac catheterization and angiography with exposure to radiation and iodinated contrast media, which has been shown to be more nephrotoxic than gadoliniumbased agents.8 Hence, a decision to use high-dose (0.4 to 0.5 mmol/kg) gadolinium was made to maximize the signal-tonoise ratio and image quality. The safety of high-dose gadolinium-based contrast has been shown by several investigators.9 –21 Most of the subjects were closely monitored in the intensive care unit after MRA, and in this small cohort, no adverse effects of the higher dose were noted. However, larger prospective dose-finding studies are needed to determine the optimal dose of gadolinium in this patient popu-
720
The American Journal of Cardiology (www.AJConline.org)
Figure 3. Oblique coronal maximum-intensity projections from a newborn infant with functionally single ventricle showing the right (A) and left (B) pulmonary arteries and branches to the upper (U), middle (M), and lower (L) lobes.
Figure 4. Maximum-intensity projections (MIPs) from a newborn infant with situs inversus totalis. A coronal MIP (A) shows dextrocardia, left-sided liver, left-sided superior vena cava (arrow), right-sided innominate vein (arrowhead), hypoplastic ascending aorta (AO), normal-sized main pulmonary artery (MPA), and a large patent ductus arteriosus (PDA). A sagittal MIP (B) shows a hypoplastic aortic arch (arrowhead) and a large PDA.
lation. Until such data are available, the dose used should be individualized to balance the risk of using a higher dose against the risk for requiring cardiac catheterization if imaging is suboptimal with a lower dose. Because of the retrospective nature of this study, a direct comparison of the findings on MRA with other reference techniques was possible only in a subset of subjects in whom no discrepancies in diagnosis were noted on surgical inspection (n ⫽ 25) or on cardiac catheterization (n ⫽ 8). Larger prospective studies are required to evaluate the accuracy of MRA compared with reference techniques.
This study was limited by its retrospective design, relatively small sample size, and heterogenous patient population. The image quality grading score was semiquantitative and hence somewhat subjective in nature. Larger prospective studies comparing MRA directly with reference techniques such as cardiac catheterization will help determine the accuracy of this technique in specific diagnostic subgroups. 1. Greil GF, Powell AJ, Gildein HP, Geva T. Gadolinium-enhanced three-dimensional magnetic resonance angiography of pulmonary and systemic venous anomalies. J Am Coll Cardiol 2002;39:335–341.
Congenital Heart Disease/MRA of Heart Disease in Infancy 2. Geva T, Greil GF, Marshall AC, Landzberg M, Powell AJ. Gadoliniumenhanced 3-dimensional magnetic resonance angiography of pulmonary blood supply in patients with complex pulmonary stenosis or atresia: comparison with X-ray angiography. Circulation 2002;106:473– 478. 3. Prasad SK, Soukias N, Hornung T, Khan M, Pennell DJ, Gatzoulis MA, Mohiaddin RH. Role of magnetic resonance angiography in the diagnosis of major aortopulmonary collateral arteries and partial anomalous pulmonary venous drainage. Circulation 2004;109:207– 214. 4. Masui T, Seelos KC, Kersting-Sommerhoff BA, Higgins CB. Abnormalities of the pulmonary veins: evaluation with MR imaging and comparison with cardiac angiography and echocardiography. Radiology 1991;181:645– 649. 5. Valsangiacomo ER, Levasseur S, McCrindle BW, MacDonald C, Smallhorn JF, Yoo SJ. Contrast-enhanced MR angiography of pulmonary venous abnormalities in children. Pediatr Radiol 2003;33:92–98. 6. Khan MA, Torres AJ, Printz BF, Prakash A. usefulness of magnetic resonance angiography for diagnosis of scimitar syndrome in early infancy. Am J Cardiol 2006;96:1313–1316. 7. Tsai-Goodman B, Geva T, Odegard KC, Sena LM, Powell AJ. Clinical role, accuracy, and technical aspects of cardiovascular magnetic resonance imaging in infants. Am J Cardiol 2004;94:69 –74. 8. Prince MR, Arnoldus C, Frisoli JK. Nephrotoxicity of high-dose gadolinium compared with iodinated contrast. J Magn Reson Imaging 1996;6:162–166. 9. Shellock FG, Kanal E. Safety of magnetic resonance imaging contrast agents. J Magn Reson Imaging 1999;10:477– 484. 10. Niendorf H, Dinger J, Haustein J, Cornelius I, Alhassan A, Clauss W. Tolerance data of Gd-DTPA: a review. Eur J Radiol 1991;13:15–20. 11. Oksendal A, Hals P. Biodistribution and toxicity of MR imaging contrast media. J Magn Reson Imaging 1993;3:157–165. 12. Niendorf H, Haustein J, Cornelius I, Alhassan A, Claus W. Safety of gadolinium-DTPA: extended clinical experience. Magn Reson Med 1991;22:222–228.
721
13. Yoshikawa K, Davies A. Safety of ProHance in special populations. Eur Radiol 1997:7(suppl):246 –250. 14. Frank J, Choyke P, Girton M, Morrison P, Diggs R, Skinner M. Gadopentate dimeglumine clearance in renal insufficiency in rabbits. Invest Radiol 1990;25:1212–1216. 15. Jau P, Bonnet JL, Joly P, Barth P, Habib G, Djiane P, Bory M, Bernard PJ. Study of early myocardial infarction by nuclear magnetic resonance imaging with gadolinium Dota injection. Arch Mal Coeur Vaiss 1991;84:195–200. 16. Remy-Jardin M, Bahepar J, Lafitte JJ, Dequiedt P, Ertzbischoff O, Bruzzi J, Delannoy-Deken V, Duhamel A, Remy J. Multi-detector row CT angiography of pulmonary circulation with gadolinium-based contrast agent: prospective evaluation in 60 patients. Radiology 2006;238: 1022–1035. 17. Remy-Jardin M, Dequiedt P, Etzbischoff O, Tillie-Leblond I, Bruzzi J, Duhamel A, Remy J. Safety and effictiveness of gadolinium-enhanced multi-detector row spiral CT angiography of the chest: preliminary results in 37 patients with contraindications to iodinated contrast agents. Radiology 2005;235:819 – 826. 18. Haustein J, Laniado M, Niendorf HP, Louton T, Beck W, Planitzer J, Schoffel M, Reiser M, Kaiser W, Schomer W. Triple-dose versus standard-dose gadopentetate dimeglumine: a randomized study in 199 patients. Radiology 1993;186:855– 860. 19. Haussler MD, Rummeny EJ, Raufhake C, Blasius S, Lindner N, Daldrup HE, Reimer P, Peters PE. High dosage Gd-DTPA-BMA (Gadodiamid) administration in diagnosis and therapeutic monitoring of malignant bone tumors. Radiology 1996;32:148 –152. 20. Runge RM, Kirsch JE, Thomas GS. High-dose Applications of gadolinium chelates in magnetic resonance imaging. Magn Reson Med 1991;22:358 –363. 21. Rosioreanu A, Hon M, Imbriano L, Mueller R, Katz DS. High-dose intravenous gadolinium for renal computed tomographic angiography. J Vasc Interv Radiol 2004;15:517–519.
Three-dimensional contrast-enhanced magnetic resonance angiography (MRA) is a noninvasive technique that has been shown to be accurate in the delineation of vascular anatomy in adults and older children and may be used instead of x-ray angiography.1–5 However, the routine use of this technique in young infants (aged ⬍3 months) has been limited by technical problems, including low image signalto-noise ratio, artifacts related to respiratory motion, hypothermia, bolus timing, and the lack of adequately sized receiver coils. The effects of these technical limitations on the quality of the visualization of individual thoracic vessels and on the diagnostic accuracy of MRA in this age group are not known. To define the role of MRA in the evaluation of these infants, it is important to systematically assess the quality of the visualization and diagnostic accuracy of MRA related to each thoracic vascular structure of interest in this patient population. We evaluated the image quality of the pulmonary arterial bed and other extracardiac thoracic vessels by MRA in newborns and young infants with congenital heart disease (CHD).
a
Division of Pediatric Cardiology and bDepartment of Radiology, Columbia University College of Physicians and Surgeons; and cDivision of Pediatric Cardiology, Mount Sinai School of Medicine, New York, New York. Manuscript received January 5, 2007; revised manuscript received and accepted March 7, 2007. *Corresponding author: Tel: 212-305-8261; fax: 212-305-4429. E-mail address: [email protected] (A. Prakash). 0002-9149/07/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2007.03.090
Methods Study population: All infants aged ⬍3 months with CHD who underwent contrast-enhanced MRA at Morgan Stanley Children’s Hospital of New York Presbyterian from March 2004 to February 2005 were included in this retrospective study. The primary diagnosis for each subject is listed in Table 1. The Columbia University Institutional Review Board gave permission for the review of existing data. The decision to perform MRA and/or cardiac catheterization was made by the patient’s cardiologist on a caseby-case basis. Magnetic resonance angiographic protocol: Magnetic resonance angiographic studies were performed using a 1.5-T whole-body scanner (Signa EXCITE; General Electric, Milwaukee, Wisconsin), using a commercially available 4-element head-imaging coil. All infants were sedated, paralyzed, and mechanically ventilated during the examinations. MRA was performed in the coronal orientation using the following imaging parameters: field of view 200 mm, matrix 192 (phase) and 256 (frequency), slice thickness 2 mm, echo time 1.1 ms, repetition time 3.4 to 3.7 ms, flip angle 40°, receiver band width 31.3 kHz, number of signal averages 1, sequential phase ordering, number of dynamics 3, and acquisition time 25 to 35 s/dynamic. Gadopentate dimeglumine (Magnevist; Berlex Laboratories, Seattle, Washington) (0.4 to 0.5 mmol/kg) followed by a normal saline flush (1 cm3/kg) was manually injected in a peripheral or central vein in the upper or lower extremity at a rate of approximately 0.5 to 1 cm3/s. After the suspension of www.AJConline.org
716
The American Journal of Cardiology (www.AJConline.org)
Table 1 Primary cardiac diagnoses
were graded separately. The aortopulmonary collaterals were graded as a group in each subject.
Diagnosis
n
Single ventricle physiology Pulmonary venous anomaly Scimitar syndrome Tetralogy of Fallot, pulmonary atresia Left pulmonary artery sling Situs inversus, coarctation of aorta Truncus arteriosus with isolated left pulmonary artery Tetralogy of Fallot, absent pulmonary valve
10 4 4 5 2 1 1 1
Table 2 Image quality grading scale Grade 0 1 2 3 A
Description Structure not seen due to poor image quality Structure seen with indistinct margins, image quality insufficient for diagnosis Structure seen with indistinct margins, image quality sufficient for diagnosis but not for measurement Structure seen with distinct margins, image quality sufficient for diagnosis and measurement Structure absent
mechanical ventilation, imaging was initiated approximately 5 to 10 seconds after the start of the injection, taking into consideration the site of the intravenous line and the length of tubing. Three sets of images (dynamics) were obtained, with ventilator breaths delivered for approximately 5 seconds between each period of apnea. The number of dynamics obtained during each period of apnea was based on the infant’s ability to tolerate apnea and varied from 1 to 2. Oxygen saturation, blood pressure, respiratory rate, and heart rate were monitored noninvasively throughout the procedure. Magnetic resonance angiographic image analysis for image quality: Magnetic resonance angiographic images were electronically deidentified and transferred to a blinded observer (JCN) who was an expert in the magnetic resonance imaging of CHD at a different institution (Mount Sinai Medical Center). This blinded observer performed all image analyses using a commercially available computer workstation and software (Advantage Windows 4.0; General Electric). Cardiac and extracardiac vascular anatomy was defined and graded by constructing user-defined subvolume maximal-intensity projections and multiplanar reformatted images from the 3-dimensional data sets. The quality of the visualization of the pulmonary arterial bed, pulmonary and systemic veins, thoracic aorta and its branches, aortopulmonary collaterals, patent ductus arteriosus, and the sidedness of the abdominal viscera was assessed semiquantitatively using the grading scale listed in Table 2. The pulmonary artery branch to each lung and further branches to each lobe of the lung were graded separately, and the second-generation branches for each lung were graded as a group. The pulmonary vein branches draining each lung lobe were graded separately. The subclavian, carotid, and internal mammary arteries on each side
Comparison with findings of x-ray angiography and surgical inspection: The specific diagnostic questions at referral for each magnetic resonance angiographic scan were determined from the official magnetic resonance angiographic report. Subjects in whom specific confirmation of the diagnostic question by x-ray angiography or surgical inspection (performed within 3 months of MRA) was available were identified. For these subjects, the magnetic resonance angiographic diagnoses for each anatomic segment in question were recorded by an observer who is an expert in the magnetic resonance imaging of CHD (AP). For all subjects, the magnetic resonance angiographic diagnosis of the anatomic segment in question preceded the catheterization or surgical procedure. X-ray angiographic images for these subjects were reviewed by an expert in congenital cardiac catheterization (AJT), and all vascular diagnoses were recorded. The findings of surgical inspection were obtained from the official operative reports. The accuracy of magnetic resonance angiographic diagnosis related to the diagnostic question at referral was thus determined. In addition, official magnetic resonance angiographic, x-ray angiographic, and surgical reports were carefully searched for any other diagnostic discrepancies, which were recorded. Results Subjects and diagnostic questions: Twenty-nine contrast-enhanced magnetic resonance angiographic scans were performed on 28 infants aged ⬍3 months (median 6 days, range 1 to 90 days; median weight 3 kg, range 2.1 to 3.9) during the 1-year study period. The results of MRA in the 4 patients with scimitar syndrome have been reported previously.6 All subjects had undergone echocardiography before MRA. The diagnostic questions at referral for MRA are listed in Table 3. MRA and sedation: All subjects tolerated the suspension of respiration for a median of 25 seconds (range 20 to 48) without significant desaturation or change in heart rate. The median total duration of scanning was 9 minutes (range 3 to 46). There were no complications. Magnetic resonance angiographic image quality: Image quality grades for analyzed structures are listed in Table 4. Overall magnetic resonance angiographic image quality was excellent in every subject. The mean image quality grade was ⬎2 for all structures except the second-generation pulmonary arterial branches, for which it was 2. The image quality grade was ⱖ2 (adequate for diagnosis) for every structure in every subject except for the secondgeneration pulmonary arterial branches. The image quality grade was ⬍2 (inadequate for diagnosis) for the secondgeneration pulmonary arterial branches in only 1 subject with tetralogy of Fallot with pulmonary atresia and severely hypoplastic native pulmonary arteries. Examples of maximum intensity projections reconstructed from magnetic resonance angiographic data sets are shown in Figures 1 to 4. Comparison with findings of x-ray angiography and surgical inspection: The results of x-ray angiography (n ⫽ 8) and/or surgical inspection (n ⫽ 25) were available for 23
Congenital Heart Disease/MRA of Heart Disease in Infancy
717
Table 3 Accuracy of magnetic resonance angiographic assessment of referral diagnostic questions Anatomic Segment for Which MRA Evaluation Was Requested
No. of MRA Scans
Diagnostic Confirmation of Anatomic Segment by Angiography/Surgery Available (n)
No. of Scans in Which MRA Diagnosis Was Concordant With Angiography/Surgery (n)
20 12 13 7 2 3 1
16 11 9 6 2 2 1
16 11 9 6 2 2 1
Pulmonary artery anatomy Pulmonary venous anatomy Aortopulmonary collaterals Anatomy of aortic arch, its branches, or ductus arteriosus Surgical psuedoaneurysm Systemic venous anatomy Airway compression
Table 4 Image quality grades Structure Analyzed
PAs Main PA Proximal PA branches Lobar PA branches Second-generation PA branches PVs Lobar pulmonary veins Anomalous PV confluence Anomalous vertical PV Aortopulmonary collateral arteries Superior vena cavae Right sided Left sided Inferior vena cava Aorta and branches Ascending aorta Transverse aortic arch Isthmus Thoracic descending aorta Carotid and subclavian arteries Internal mammary arteries Patent ductus arteriosus Visceral sidedness Stomach Liver Spleen
Image Quality Grade
No. of Scans in Which Structure Was Mean ⫾ SD Range Present 3⫾0 3⫾0 2.8 ⫾ 0.6 2 ⫾ 0.6
3–3 3–3 2–3 0–3
26 29 29 29
2.9 ⫾ 0.1 3⫾0 3⫾0 2.7 ⫾ 0.5
2–3 3–3 3–3 2–3
29 3 4 6
3⫾0 3⫾0 3⫾0
2–3 2–3 3–3
26 10 28
3⫾0 3⫾0 3⫾0 3⫾0 3⫾0 2.9 ⫾ 0.3 2.9 ⫾ 0.3
3–3 3–3 3–3 3–3 3–3 2–3 2–3
29 27 27 29 29 29 13
3⫾0 3⫾0 3⫾0
3–3 3–3 3–3
29 29 24
PA ⫽ pulmonary artery; PV ⫽ pulmonary vein.
and 28 subjects, respectively. Comparisons of the diagnostic findings of MRA with those of x-ray angiography and surgical inspection, as related to the diagnostic questions at referral, are listed in Table 3. The diagnostic questions at referral were accurately answered by MRA in each subject. No discrepancies were noted between the official magnetic resonance angiographic, x-ray angiographic, and operative reports. For the aortopulmonary collaterals, the number and sites of origin of all collaterals were accurately assessed by MRA. The distribution of branches from the collaterals to individual lung lobes was accurately assessed by MRA in 7 of 9 subjects; in 2 of 9 subjects, there were minor discrepancies between MRA and x-ray angiography in the delineation of the lobar lung distribution of the aortopulmonary
collaterals. The presence of “dual” blood supply to individual lung segments via collaterals and native pulmonary arteries was not assessed by MRA. Indications for cardiac catheterization: Although the diagnostic questions at referral were accurately answered by MRA, in 6 of 28 subjects, additional cardiac catheterization was performed before surgery. The purpose of catheterization was diagnostic in 3 subjects (to confirm the anatomy and lung-segment distribution of the aortopulmonary collaterals in 2 subjects with tetralogy of Fallot with pulmonary atresia and to measure pulmonary artery pressure in 1 subject with scimitar syndrome, ventricular septal defect, and left pulmonary artery sling). In the other 3 subjects, catheterization was performed for interventional procedures (stenting of pulmonary vein stenosis in 1 subject, decompression of the left atrium in 1 subject with hypoplastic left heart syndrome, and embolization of an aortopulmonary collateral in 1 subject with scimitar syndrome). Adverse effects: No immediate adverse effects were noted after the injection of contrast medium. Clinical and laboratory data for a period of ⱖ72 hours were available after 25 of 29 scans in subjects who remained admitted to the hospital. None of these subjects demonstrated an increase in serum creatinine, abnormal hepatic function, or other adverse effects during this period. Three of 29 subjects were discharged home soon after the magnetic resonance imaging scans, and although laboratory data were unavailable, none had a clinical adverse event on follow-up. In 1 subject, the magnetic resonance imaging scan was performed emergently before cardiac surgery. This patient died in the operating room. Discussion The present study demonstrates the utility of MRA in young infants with complex CHD in whom echocardiographic evaluation is incomplete or inconclusive. Because echocardiography provides excellent visualization of intracardiac anomalies, the referral question was frequently related to the extracardiac thoracic vasculature. MRA is especially suited for the evaluation of these anomalies, especially when they occur in combination with each other. After a single intravenous injection of contrast medium, a 3-dimensional data set of the entire vasculature can be obtained rapidly. The present report is the first to systematically study the quality of the visualization of individual extracardiac vascular structures in a group of infants aged ⬍3 months with complex
718
The American Journal of Cardiology (www.AJConline.org)
Figure 1. Maximum-intensity projections (MIPs) from a newborn infant with heterotaxy (asplenia) syndrome, functionally single ventricle, double-outlet right ventricle, pulmonary stenosis, and total anomalous pulmonary venous return showing (A) right superior vena cava (RSVC) and left superior vena cava (LSVC) with a small bridging vein; (B) levocardia with a midline transverse liver; and (C) confluence of right upper (RU), right lower (RL), left upper (LU), and left lower (LL) pulmonary veins behind the left atrium draining superiorly via a vertical vein (VV). A sagittal MIP (D) shows drainage of the VV into the LSVC.
CHD. Image quality was excellent for the diagnosis and measurement of anomalies involving the main, proximal branch, and lobar pulmonary arteries; the systemic and pulmonary veins; the aorta and its branches; and the aortopulmonary collateral arteries and for the determination of abdominal visceral situs. Image quality grade for the second-generation pulmonary arteries was lower on average, and they were not well visualized in 1 subject with tetralogy of Fallot, pulmonary atresia, and hypoplastic native pulmonary arteries. The acquisition parameters used in this series were chosen to provide adequate spatial resolution while maintaining acceptable signal-to-noise characteristics on the basis of our institutional experience, the scanning platform used, and the work of others.7 The field of view and imaging matrix were adjusted to maximize resolution while balancing signal-tonoise characteristics. Parallel-processing techniques such as sensitivity encoding were purposefully not used to maximize the signal-to-noise ratio and because relatively long
periods of apnea are well tolerated by intubated infants. The contrast was injected manually with the “best-estimate” method used for the timing of imaging. Although real-time imaging sequences are now commercially available to allow the timing of contrast bolus, we prefer the best-estimate method in small infants because of their rapid heart rates and relatively fast circulation times. The manual injection of contrast has the advantage of allowing direct monitoring of the subject and communication with the respiratory therapist during the suspension of respiration. In our experience, the site of injection (upper or lower extremity) and the type of venous line (central or peripheral) do not affect the quality of imaging significantly, although this was not evaluated critically in this study. We also believe that a rapid injection rate (0.5 to 1 cm3/s) is important to allow scanning during the “first pass” of the contrast medium because of the fast circulation time in small infants, especially in the presence of intracardiac shunts. The elimination of respiratory
Congenital Heart Disease/MRA of Heart Disease in Infancy
719
Figure 2. Maximum-intensity projections (MIPs) from a newborn infant with hypoplastic left-sided cardiac syndrome after stage 1 palliation with a right ventricle–to–pulmonary artery conduit. Axial (A), sagittal (B), and coronal (C) MIPs show a large psuedoaneurysm (arrow) anterior to the right ventricle–to–pulmonary artery conduit (arrowhead). The reconstructed aortic arch and the aortopulmonary anastomosis are also well seen in a sagittal plane (D). LPA ⫽ left pulmonary artery; RPA ⫽ right pulmonary artery.
motion is important in this group of patients because even slight blurring can make the visualization of small structures difficult in these small-sized subjects. However, this requires endotracheal intubation and paralysis with close monitoring by either an anesthesiologist or an intensivist. These resources may not be available at all institutions, and hence the decision to use intubation and paralysis should be individualized. The median total scanning duration was 9 minutes, showing that in most infants, the referral question can be answered rapidly using MRA. Additional functional assessment, including the quantification of ventricular function and the measurement of blood flow, can be performed using magnetic resonance imaging. However, in most subjects in this series, these data were provided by echocardiography alone or were not thought to be clinically relevant. The optimal dose for MRA of complex CHD in infants aged ⬍3 months is unknown. Dose-finding studies are very difficult to perform in this patient population because of ethical and logistic reasons. This is true for most drugs used
in infants. Accurate diagnoses of the often complex cardiac and vascular anomalies in this patient group require imaging quality of an order that is often limited by the inherently low signal-to-noise ratio related to their small body size. To an extent, the signal-to-noise ratio can be improved with the use of a higher dose of contrast. In this reported cohort, failure to obtain diagnostic images on MRA would have necessitated cardiac catheterization and angiography with exposure to radiation and iodinated contrast media, which has been shown to be more nephrotoxic than gadoliniumbased agents.8 Hence, a decision to use high-dose (0.4 to 0.5 mmol/kg) gadolinium was made to maximize the signal-tonoise ratio and image quality. The safety of high-dose gadolinium-based contrast has been shown by several investigators.9 –21 Most of the subjects were closely monitored in the intensive care unit after MRA, and in this small cohort, no adverse effects of the higher dose were noted. However, larger prospective dose-finding studies are needed to determine the optimal dose of gadolinium in this patient popu-
720
The American Journal of Cardiology (www.AJConline.org)
Figure 3. Oblique coronal maximum-intensity projections from a newborn infant with functionally single ventricle showing the right (A) and left (B) pulmonary arteries and branches to the upper (U), middle (M), and lower (L) lobes.
Figure 4. Maximum-intensity projections (MIPs) from a newborn infant with situs inversus totalis. A coronal MIP (A) shows dextrocardia, left-sided liver, left-sided superior vena cava (arrow), right-sided innominate vein (arrowhead), hypoplastic ascending aorta (AO), normal-sized main pulmonary artery (MPA), and a large patent ductus arteriosus (PDA). A sagittal MIP (B) shows a hypoplastic aortic arch (arrowhead) and a large PDA.
lation. Until such data are available, the dose used should be individualized to balance the risk of using a higher dose against the risk for requiring cardiac catheterization if imaging is suboptimal with a lower dose. Because of the retrospective nature of this study, a direct comparison of the findings on MRA with other reference techniques was possible only in a subset of subjects in whom no discrepancies in diagnosis were noted on surgical inspection (n ⫽ 25) or on cardiac catheterization (n ⫽ 8). Larger prospective studies are required to evaluate the accuracy of MRA compared with reference techniques.
This study was limited by its retrospective design, relatively small sample size, and heterogenous patient population. The image quality grading score was semiquantitative and hence somewhat subjective in nature. Larger prospective studies comparing MRA directly with reference techniques such as cardiac catheterization will help determine the accuracy of this technique in specific diagnostic subgroups. 1. Greil GF, Powell AJ, Gildein HP, Geva T. Gadolinium-enhanced three-dimensional magnetic resonance angiography of pulmonary and systemic venous anomalies. J Am Coll Cardiol 2002;39:335–341.
Congenital Heart Disease/MRA of Heart Disease in Infancy 2. Geva T, Greil GF, Marshall AC, Landzberg M, Powell AJ. Gadoliniumenhanced 3-dimensional magnetic resonance angiography of pulmonary blood supply in patients with complex pulmonary stenosis or atresia: comparison with X-ray angiography. Circulation 2002;106:473– 478. 3. Prasad SK, Soukias N, Hornung T, Khan M, Pennell DJ, Gatzoulis MA, Mohiaddin RH. Role of magnetic resonance angiography in the diagnosis of major aortopulmonary collateral arteries and partial anomalous pulmonary venous drainage. Circulation 2004;109:207– 214. 4. Masui T, Seelos KC, Kersting-Sommerhoff BA, Higgins CB. Abnormalities of the pulmonary veins: evaluation with MR imaging and comparison with cardiac angiography and echocardiography. Radiology 1991;181:645– 649. 5. Valsangiacomo ER, Levasseur S, McCrindle BW, MacDonald C, Smallhorn JF, Yoo SJ. Contrast-enhanced MR angiography of pulmonary venous abnormalities in children. Pediatr Radiol 2003;33:92–98. 6. Khan MA, Torres AJ, Printz BF, Prakash A. usefulness of magnetic resonance angiography for diagnosis of scimitar syndrome in early infancy. Am J Cardiol 2006;96:1313–1316. 7. Tsai-Goodman B, Geva T, Odegard KC, Sena LM, Powell AJ. Clinical role, accuracy, and technical aspects of cardiovascular magnetic resonance imaging in infants. Am J Cardiol 2004;94:69 –74. 8. Prince MR, Arnoldus C, Frisoli JK. Nephrotoxicity of high-dose gadolinium compared with iodinated contrast. J Magn Reson Imaging 1996;6:162–166. 9. Shellock FG, Kanal E. Safety of magnetic resonance imaging contrast agents. J Magn Reson Imaging 1999;10:477– 484. 10. Niendorf H, Dinger J, Haustein J, Cornelius I, Alhassan A, Clauss W. Tolerance data of Gd-DTPA: a review. Eur J Radiol 1991;13:15–20. 11. Oksendal A, Hals P. Biodistribution and toxicity of MR imaging contrast media. J Magn Reson Imaging 1993;3:157–165. 12. Niendorf H, Haustein J, Cornelius I, Alhassan A, Claus W. Safety of gadolinium-DTPA: extended clinical experience. Magn Reson Med 1991;22:222–228.
721
13. Yoshikawa K, Davies A. Safety of ProHance in special populations. Eur Radiol 1997:7(suppl):246 –250. 14. Frank J, Choyke P, Girton M, Morrison P, Diggs R, Skinner M. Gadopentate dimeglumine clearance in renal insufficiency in rabbits. Invest Radiol 1990;25:1212–1216. 15. Jau P, Bonnet JL, Joly P, Barth P, Habib G, Djiane P, Bory M, Bernard PJ. Study of early myocardial infarction by nuclear magnetic resonance imaging with gadolinium Dota injection. Arch Mal Coeur Vaiss 1991;84:195–200. 16. Remy-Jardin M, Bahepar J, Lafitte JJ, Dequiedt P, Ertzbischoff O, Bruzzi J, Delannoy-Deken V, Duhamel A, Remy J. Multi-detector row CT angiography of pulmonary circulation with gadolinium-based contrast agent: prospective evaluation in 60 patients. Radiology 2006;238: 1022–1035. 17. Remy-Jardin M, Dequiedt P, Etzbischoff O, Tillie-Leblond I, Bruzzi J, Duhamel A, Remy J. Safety and effictiveness of gadolinium-enhanced multi-detector row spiral CT angiography of the chest: preliminary results in 37 patients with contraindications to iodinated contrast agents. Radiology 2005;235:819 – 826. 18. Haustein J, Laniado M, Niendorf HP, Louton T, Beck W, Planitzer J, Schoffel M, Reiser M, Kaiser W, Schomer W. Triple-dose versus standard-dose gadopentetate dimeglumine: a randomized study in 199 patients. Radiology 1993;186:855– 860. 19. Haussler MD, Rummeny EJ, Raufhake C, Blasius S, Lindner N, Daldrup HE, Reimer P, Peters PE. High dosage Gd-DTPA-BMA (Gadodiamid) administration in diagnosis and therapeutic monitoring of malignant bone tumors. Radiology 1996;32:148 –152. 20. Runge RM, Kirsch JE, Thomas GS. High-dose Applications of gadolinium chelates in magnetic resonance imaging. Magn Reson Med 1991;22:358 –363. 21. Rosioreanu A, Hon M, Imbriano L, Mueller R, Katz DS. High-dose intravenous gadolinium for renal computed tomographic angiography. J Vasc Interv Radiol 2004;15:517–519.