Anatomical and Functional Evaluation of Pulmonary Veins in Children by Magnetic Resonance Imaging

Anatomical and Functional Evaluation of Pulmonary Veins in Children by Magnetic Resonance Imaging

Journal of the American College of Cardiology © 2007 by the American College of Cardiology Foundation Published by Elsevier Inc. Vol. 49, No. 9, 2007...

7MB Sizes 124 Downloads 45 Views

Journal of the American College of Cardiology © 2007 by the American College of Cardiology Foundation Published by Elsevier Inc.

Vol. 49, No. 9, 2007 ISSN 0735-1097/07/$32.00 doi:10.1016/j.jacc.2006.09.052

Congenital Heart Disease: State-of-the-Art Paper

Anatomical and Functional Evaluation of Pulmonary Veins in Children by Magnetic Resonance Imaging Lars Grosse-Wortmann, MD,*† Abdulmajeed Al-Otay, MD,*† Hyun Woo Goo, MD,* Christopher K. Macgowan, PHD,* John G. Coles, MD,‡ Leland N. Benson, MD,† Andrew N. Redington, MD,† Shi-Joon Yoo, MD*† Toronto, Canada Pulmonary vein pathologies often present a diagnostic challenge. Among the different imaging modalities used for the evaluation of pulmonary veins, magnetic resonance is the most comprehensive in assessing anatomy and pathophysiology at the same time. Bright blood cine sequences and contrast-enhanced magnetic resonance angiography outline the course and connections of the pulmonary veins. Phase-contrast velocity mapping measures flow patterns, velocities, and volumes throughout the pulmonary circulation. This paper reviews contemporary utilization of magnetic resonance in the evaluation of pulmonary venous abnormalities in children, based on our experience over the last 5 years and on that of other investigators. We summarize how magnetic resonance imaging enhances our understanding of pulmonary vein physiology and how it can influence the diagnostic approach to children and adults with a pulmonary venous pathology, and we discuss its limitations. (J Am Coll Cardiol 2007;49:993–1002) © 2007 by the American College of Cardiology Foundation

While relatively rare, abnormalities of the pulmonary veins represent one of the more challenging areas of diagnosis and treatment in children. Although echocardiography, cardiac catheterization with angiography, computed tomography (CT), and radioisotope ventilation/perfusion scanning all play important roles in the diagnosis and functional assessment of pulmonary venous anomalies, magnetic resonance imaging (MRI) is considered best suited for pulmonary venous abnormalities for 3 reasons (1). First, MRI is able to visualize vessels regardless of their orientation; second, it provides not only the anatomical details but also hitherto unmeasurable functional information such as volumetric blood flow and temporal flow relationships; and third, it is a non-invasive procedure without radiation. While MRI has become the “one-stop-shop” study for the assessment of the pulmonary arteries in many conditions, there are only a few reports on the assessment of the pulmonary veins (2– 4). This article aims to describe the MRI approach to pulmonary venous anatomy and function and to summarize how this can enhance our understanding of the pathophysiology.

From the *Section of Cardiac Imaging, Department of Diagnostic Imaging; †Division of Cardiology, Department of Paediatrics; and the ‡Division of Cardiovascular Surgery, Department of Surgery, The Hospital for Sick Children, The University of Toronto, Toronto, Canada. Manuscript received July 18, 2006; revised manuscript received September 18, 2006, accepted September 27, 2006.

MRI Techniques Three main MRI techniques, including white-blood cine imaging, contrast-enhanced angiography (MRA), and phase-contrast velocity mapping (PC MRI) are currently used for the assessment of pulmonary venous pathologies. Magnetic resonance oximetry can be used when a pulmonary arteriovenous or an intracardiac shunt is suspected. Noncontrast white-blood imaging. White-blood imaging is usually performed by using an electrocardiographically gated segmented gradient refocused echo (GRE) sequence or a balanced steady-state free precession (SSFP) sequence. Steady state free precession sequences provide the best natural contrast between the flowing blood and surrounding structures. Therefore, and because of their short acquisition times, they are the preferred technique (5). However, because of the requirements for steady state, SSFP sequences are sensitive to field inhomogeneities that occur within the lungs or due to metallic implants. A repetition time (TR) of ⬎4 ms results in significant flow artifacts in SSFP images. Therefore, a segmented GRE sequence with multiple signal-averaging is preferred in small infants, in whom a small field of view requires a longer TR. Recently, 3-dimensional (3D) volume data of the heart and great vessels can be obtained by using a multislice multiphase 3D SSFP sequence (6). MRA. White-blood imaging sequences do not always provide sufficient spatial resolution for the more peripheral pulmonary veins, especially when the anatomy is complex. Gadolinium, used in MRA with proven safety, is a strong

994

Grosse-Wortmann et al. MRI Evaluation of Pulmonary Veins

T1-shortening agent that enables very rapid image acquisition of vascular anatomy with the signals AVM ⴝ arteriovenous from the surrounding structures malformation sufficiently suppressed (Fig. 1) CT ⴝ computed tomography (2,4). After injection, a delay of 4 GRE ⴝ gradient refocused to 8 s after the arrival of the conecho trast medium in the descending MRA ⴝ (contrastaorta allows proper visualization of enhanced) magnetic resonance angiography the pulmonary veins as well as of the pulmonary arteries and aorta MRI ⴝ magnetic resonance imaging in the same angiogram. The imaging time is ⬍20 s for a multislice PC MRI ⴝ phase-contrast (velocity mapping) 3D angiographic data set in most magnetic resonance patients. Asymmetric k-space imaging sampling and coverage of a smaller Qp/Qs ⴝ pulmonary-tovolume of interest further reduce systemic blood flow ratio scan time. The imaging time can SSFP ⴝ steady-state free be shortened by using parallel imprecession aging and undersampling of the TR ⴝ repetition time high-frequency k-space data with Tres ⴝ temporal resolution view-sharing or key-hole imaging 3D ⴝ 3-dimensional (7,8). This technique allows time%HbO2 ⴝ blood oxygen resolved 3D angiography with a saturation temporal resolution (Tres) as low as 2 s/frame (9). PC MRI. Phase-contrast velocity mapping depicts the velocity, volume, and pattern of blood flow (Fig. 2) (3,10). This sequence measures the phase shifts of moving compared with stationary nuclei. The extent of shifting is proportional to the flow velocity (11). To eliminate phase shifts caused by other factors than blood flow, a second data set is acquired with an inverted gradient and subtracted from the first. Therefore, each k-space line is sampled twice, and the Tres is TR ⫻ 2 ⫻ VPS (views or lines per segment) (11). With 1 view per segment, the best possible Tres is (2 ⫻ TR), typically ranging between 10 and 15 ms depending on the scanner. Scan time increases with higher temporal and spatial resolution, and this trade-off must always be taken into consideration. Parallel imaging reduces the imaging time without compromising the accuracy but limits the spatial resolution (12). Accurate measurement of the blood flow velocity and pattern requires high Tres, while adequate blood flow volume measurements can be achieved with relatively low Tres. In our experience, for accurate assessment of both flow volume and flow pattern, 20 to 30 real data points within a cardiac cycle represent a good compromise between acquisition time and Tres. It is important to understand that the arbitrarily chosen number of phases in a cardiac cycle does not represent the true Tres but is the number of mathematically reconstructed phases. For example, a patient who has a heart rate of 100 beats/min and therefore an R-R interval of 600 ms requires a minimum Tres of 30 ms to achieve 20 data points per cardiac cycle. Thus, if TR is 7.5 ms, the maximum allowable views per segment is 2. The imaging plane should be strictly

JACC Vol. 49, No. 9, 2007 March 6, 2007:993–1002

Abbreviations and Acronyms

Figure 1

Contrast-Enhanced Angiograms of Normal Pulmonary Veins

(A) Two-dimensional reformatted image. (B) Volume-rendered 3-dimensional image. The right lung has 4 pulmonary veins (asterisks), 3 of which form a confluence (C) as they drain into the left atrium (LA). Ao ⫽ aorta; LPA ⫽ left pulmonary artery; RPA ⫽ right pulmonary artery.

tailored perpendicular to the blood flow using the double oblique technique. Of note, flow velocities are more sensitive to angulation errors than flow volume measurements. Velocities are underestimated by a factor equal to the cos of the angle between the normal to the imaging plane and the vessel axis. The cross-sectional area is overestimated by a factor equal to the inverse of the cos of the same angle. Both contributions cancel as flow volume is the product of area and velocity. In contrast with Doppler ultrasound, PC MRI allows full freedom to adjust the imaging planes as needed for accurate analysis of flow volume and pattern. Even the pulmonary

JACC Vol. 49, No. 9, 2007 March 6, 2007:993–1002

Figure 2

Phase-Contrast Images and Flow Velocity Curve of the Right Upper PV

(A) Magnitude image showing the anatomy of the slice. The cross section of the right upper pulmonary vein (PV) is seen below the oblique section of the right pulmonary artery (RPA) and behind the superior vena cava (SVC). (B) A phase image obtained in systole shows the flow through the RPA in black and the flow through the right upper PV in white. (C) Time-velocity curves of the right pulmonary artery and right upper pulmonary vein. The pulmonary venous blood flow consists of systolic and diastolic (d) forward waves and a late diastolic reversed wave (a) on atrial contraction. The systolic wave consists of early (es) and late (ls) systolic peaks and is slightly higher than the diastolic wave. Ao ⫽ aorta; RA ⫽ right atrium.

vessels within the aerated lungs are approachable with MRI, albeit accurate quantification is difficult because of artifacts from adjacent air (13). Therefore, exact flow quantification is possible only for the vessels within the mediastinum in most patients. Flow patterns, on the other hand, can be assessed not only centrally but also in the peripheral veins. Phase-contrast MRI flow measurements are subject to an inaccuracy of ⬍5% (10,14). Flow information of the target vessel can be compared with the information from other vessels that are included in the imaging plane. This allows comparison of the timing of peaks and troughs of the flow curves of the target pulmonary vein with those of the pulmonary arterial and systemic arterial flow curves (Fig. 2) (15). Recently, real-time velocity encoding has been introduced (16). Phase-contrast MRI of the pulmonary veins is obtained without breath-hold because the venous flow pattern can change significantly with breath-holding (17). In contrast

Grosse-Wortmann et al. MRI Evaluation of Pulmonary Veins

995

with echocardiography, free breathing PC MRI incorporates multiple full respiratory cycles and, therefore, yields the average blood flow for in- and expiration. Although this nature of PC MRI precludes assessment of the effects of respiration on pulmonary venous flow pattern, the averaged data from PC MRI might be more representative of the hemodynamics. The normal pulmonary venous flow curve consists of 2 forward flow waves during systole and early diastole followed by a short reversed flow wave when the left atrium contracts in late diastole (Fig. 2C). The systolic wave usually has 2 peaks. The first peak of the systolic venous flow wave is driven by a suction effect of the left atrium with a fall in atrial pressure after atrial contraction, whereas the second peak is a result of the net forward blood flow through the lungs generated by right ventricular systole (18). In our previous study, the mean flow velocities of the systolic and diastolic peaks of the normal pulmonary veins were 51 ⫾ 16 cm/s and 47 ⫾ 11 cm/s, respectively, with a ratio between the 2 being 1.1 ⫾ 0.3 (13). Oximetry. Measurement of the blood oxygen saturation (%HbO2) in different parts of the systemic and pulmonary circulation is used to estimate shunt volumes, cardiac output, and perfusion of peripheral organs. Magnetic resonance imaging oximetry is based on the mathematical relationship between the T2 relaxation time of blood and the %HbO2. This relationship can be expressed by the equation: 1/T2 ⫽ 1/ T2O ⫹ K {(1 ⫺ %O2)/100}2, where T2O is the T2 signal decay of fully oxygenated blood and K is a constant (19). Thus, the %HbO2 can be estimated by a T2-weighted imaging sequence after a calibration to determine the T2O and K in the subject (20). Recent studies found that both T2O and the constant K can be estimated if the patient’s hematocrit and fibrinogen concentrations are known (21). Magnetic resonance imaging oximetry can be applied for detection of intra- or extra-cardiac shunt, such as pulmonary arteriovenous fistulous communication, and is expected to become more prevalent with a general trend for avoiding cardiac catheterization. Clinical Applications Anomalous pulmonary venous connections. Isolated total and partial anomalous pulmonary venous connections account for 1.5% and 0.5%, respectively, of congenital heart disease (22). The evaluation requires visualization of all pulmonary veins and other pulmonary-to-systemic venous connections. We believe MRI to be the method of choice because additional hemodynamic information can be obtained (1,2). A complete study should, therefore, include measurement of the volume and blood flow velocity of the abnormal venous drainage, pulmonary-to-systemic blood flow ratio (Qp/Qs), description of the secondary effects on pulmonary arterial size and blood flow, and ventricular size and function (Fig. 3) (23,24). In particular, MRI plays an indispensable role in the evaluation of scimitar syndrome, as

996

Figure 3

Grosse-Wortmann et al. MRI Evaluation of Pulmonary Veins

JACC Vol. 49, No. 9, 2007 March 6, 2007:993–1002

Partial Anomalous Pulmonary Venous Connection of the RUPV to the SVC in a 3-Year-Old Girl

(A) A contrast-enhanced 3-dimensional angiogram seen from posterior shows the anomalous connection of the right upper pulmonary vein (RUPV) to the superior vena cava (SVC). The small right lower pulmonary vein (RLPV) connects normally to the left atrium (LA). (B) Time-flow volume curves of the right upper and left upper pulmonary veins and right and left pulmonary arteries. The flow pattern of the right upper pulmonary vein is different from that of the left upper pulmonary vein. There is continuous flow without discernable systolic and diastolic waves in the anomalous right upper pulmonary vein. The right pulmonary artery shows a less resistant flow pattern with the systolic wave reaching its peak later and showing slower deceleration as compared with the left pulmonary artery. The blood flow ratio between the right and left lung is 56%:44%. The pulmonary-to-systemic blood flow ratio was 2.2, and the patient required surgical repair of the anomalous connection. Ao ⫽ aorta; RA ⫽ right atrium; RPA ⫽ right pulmonary artery.

JACC Vol. 49, No. 9, 2007 March 6, 2007:993–1002

it not only provides crucial information regarding the abnormal pulmonary venous drainage but also the anatomical and hemodynamic details of the associated abnormalities, including right lung hypoplasia, aberrant systemic arterial blood supply to the right lower lung, and blood flow distribution between the lungs (Fig. 4) (25).

Figure 4

Scimitar Syndrome in a 3-Month-Old Boy

Contrast-enhanced angiograms, reformatted in slanted coronal planes, show the scimitar vein draining most of the right lung to the inferior vena cava (IVC). (A) The scimitar vein shows severe stenosis as it connects to the IVC. Note the high signal intensity flow in the IVC. (B) A small right lower pulmonary vein (RLPV) and the left pulmonary veins (LUPV and LLPV) have normal connections to the left atrium (LA). The LUPV is stenosed. A small aberrant arterial branch arises from the abdominal descending aorta (Ao) to supply the posterior basal part of the right lower lobe. The right lung volume is smaller than the left lung volume. In this case, the right pulmonary artery (RPA) is not hypoplastic. LLPV ⫽ left lower pulmonary vein; LUPV ⫽ left upper pulmonary vein; PA ⫽ main pulmonary artery; RA ⫽ right atrium.

Grosse-Wortmann et al. MRI Evaluation of Pulmonary Veins

997

Pulmonary vein stenosis. Pulmonary vein stenosis may occur as an isolated lesion or secondary to cardiovascular surgery or interventional therapies (26). When pulmonary vein stenosis is suspected, MRA is performed as the initial MRI sequence (Fig. 5). In a study by Valsangiacomo et al. (2), MRA visualized the central two-thirds of the pulmonary veins in 99%, while echocardiography was able to identify the pulmonary veins in only 89% (2). Contrastenhanced magnetic resonance angiography visualizes not only the stenotic lesion but also the collateral venous channels to the unobstructed pulmonary veins (Fig. 6). The significance of the stenosis can be evaluated by measuring flow in the pulmonary veins and arteries in the affected and unaffected lungs. Ideally, flow data are acquired from both upstream and downstream sides of the stenosis. Typically, the obstructed pulmonary vein shows loss of phasic blood flow pattern with an increased flow velocity downstream and a decreased velocity upstream from the stenosis (Figs. 5B and 5D). When the stenosis is at the veno-atrial junction, it is difficult to target the downstream jet as its direction within the cavity of the atrium is unpredictable. In these cases, the pre-stenotic segment is easier to target and more reproducible than the downstream post-stenotic jet (3). Changes in caliber, and blood flow pattern, and velocity, however, are affected by additional factors. For example, accelerated or turbulent flow in a pulmonary vein does not necessarily mean that there is upstream stenosis (3). Valsangiacomo et al. (2) showed that the echocardiographic suspicion of pulmonary vein stenosis based on a turbulent Doppler signal was dismissed on the basis of MRA in nearly 50% of the study cases. On the other hand, vessel caliber and blood flow profile of the stenosed pulmonary vein do not necessarily reflect the real severity of the stenosis, because redistribution of blood flow to the unaffected lung areas results in reduced blood flow through and reduced caliber of the affected pulmonary vein (27). Roman et al. (27) demonstrated that unilateral pulmonary vein stenosis was associated with reduced systolic forward flow and diastolic flow reversal in the ipsilateral branch pulmonary artery (Figs. 5B and 5C). The contralateral branch pulmonary artery showed increased systolic flow and continuous forward flow in diastole. Blood flowing forward in the branch pulmonary artery of the unaffected lung during diastole was considered to originate from the reversed flow in the branch pulmonary artery on the side of the stenosed vein. Furthermore, the cross-sectional area ratio of the right and left branch pulmonary arteries correlated well with the ratio of net forward flow through the pulmonary arteries, suggesting that the hemodynamic effect of unilateral pulmonary vein stenosis is reflected in the caliber change in the branch pulmonary arteries. Blood flow redistribution can also be observed within the lung when 1 pulmonary vein is stenosed while the other veins are not. Phase-contrast MRI of the pulmonary arteries may show the flow patterns of secondary pulmonary hypertension that include an early systolic peak with reduced peak velocity, early deceleration

998

Figure 5

Grosse-Wortmann et al. MRI Evaluation of Pulmonary Veins

JACC Vol. 49, No. 9, 2007 March 6, 2007:993–1002

Congenital Stenosis of the Left Pulmonary Veins in an 8-Month-Old Girl

(A) Contrast-enhanced angiograms show severe stenosis of the left lower pulmonary vein (LLPV) and mild stenosis of the left upper pulmonary vein (LUPV). Both veins are slightly smaller than the right pulmonary veins (RUPV and RLPV). The left pulmonary artery (LPA) is smaller than the right pulmonary artery (RPA) due to diversion of blood flow to the right lung. (B) Time-velocity curves of the pulmonary veins and pulmonary arteries using phase-contrast imaging. The left pulmonary venous flow data were obtained from just upstream of the stenosis. The left pulmonary veins demonstrate reduced velocity and loss of phasic changes. The LPA has reduced systolic blood flow and a premature systolic peak. The blood flow ratio between the right and left lungs is 73%:27%. (C) Contrast-enhanced angiograms obtained 8 months after the surgical repair of the stenosis show residual tight stenosis of the LUPV and less severe narrowing of the LLPV. Both vessels are much smaller than the right pulmonary veins. The right and left pulmonary arteries remain discrepant in size. (D) Flow curves 8 months after surgery. Left lower pulmonary venous flow velocity is markedly reduced. Left pulmonary arterial systolic flow is further reduced, and there is reversed flow (shaded) in diastole. There is almost no net blood flow to the left lung. LA ⫽ left atrium; MPA ⫽ main pulmonary artery; RLPV ⫽ right lower pulmonary vein; RUPV ⫽ right upper pulmonary vein.

Grosse-Wortmann et al. MRI Evaluation of Pulmonary Veins

JACC Vol. 49, No. 9, 2007 March 6, 2007:993–1002

Figure 6

999

Postoperative Pulmonary Vein Obstruction With Collateral Drainage to the Unobstructed Ipsilateral Pulmonary Vein in a 6-Year-Old Boy

Contrast-enhanced magnetic resonance angiograms, reformatted in slanted coronal planes, reveal complete occlusion of the left upper (LUPV) and left lower (LLPV) pulmonary veins. (A) The left middle pulmonary vein (LMPV) has an unobstructed connection to the left atrium (LA). (B) Collateral channels (arrows) are seen between the peripheral branches of the unobstructed LMPV and the branches of the obstructed LUPVs and LLPVs. Ao ⫽ aorta; RA ⫽ right atrium; RLPV ⫽ right lower pulmonary vein; RPA ⫽ right pulmonary artery.

of the systolic flow curve, and multiple peaks in the deceleration phase of the systolic curve (28). Pulmonary venous flow pattern after pulmonary vein surgery. The flow pattern in unobstructed pulmonary veins after surgical repair of pulmonary venous pathology differs from the normal pattern (3). Repaired veins usually show a greater diastolic flow component than normal veins, and often contain an abnormal downward deflection wave in the early systolic phase (Fig. 7). Both features are considered to reflect impaired compliance of the left atrium either related to scar formation at the site of atrial incision or to the introduction of graft material into the atrial wall. The less compliant left atrium restricts pulmonary venous return during ventricular systole, and, therefore, a greater than normal portion of pulmonary venous drainage occurs when the mitral valve is open in early diastole. The downward deflection of the flow curve in early systole is considered to be due to a transient pressure increase in the non-compliant left atrium at the time of mitral valve closure. Quantification of pulmonary venous flow volume. In order to measure the pulmonary venous return with PC MRI, each pulmonary vein is imaged just proximal to the veno-atrial junction. The sum of the return (i.e., the total pulmonary venous blood flow) is usually slightly more than the total pulmonary arterial blood flow because of the addition of bronchial arterial blood flow, which is, at least in part, drained through the pulmonary veins. This difference in flow volume becomes more pronounced when there are additional pathological sources of blood flow to the lungs.

The lungs recruit additional sources of blood supply after creation of a cavopulmonary anastomosis, in which systemic arterial collateral supply to the lungs is seen in the majority of patients (29). As the systemic collateral arterial circulation in this setting is an inefficient shunt of oxygenated blood ultimately causing volume overload, large collateral arteries are percutaneously occluded. However, few objective criteria are available defining the necessity of and the

Figure 7

Time-Velocity Curves in a 6-Year-Old Girl With Repaired Anomalous Pulmonary Venous Connection

The unobstructed right upper pulmonary vein demonstrates the usual postoperative changes in flow pattern including a higher diastolic wave (d) than the early (es) and late (ls) systolic peaks. A reversed flow peak (asterisk) is seen in early systole as the mitral valve closes toward the stiff left atrium.

1000

Grosse-Wortmann et al. MRI Evaluation of Pulmonary Veins

benefit from this procedure. In this context, the systemic collateral arterial supply to each lung can be quantified by PC MRI by subtracting the flow volume through the branch pulmonary artery from the volume of pulmonary venous drainage (Fig. 8) (30). The amount of pulmonary venous return can be used as a surrogate for the pulmonary arterial blood flow volume when direct measurement of the arterial blood flow is impossible. These conditions include previous stent placement and significantly turbulent flow in the pulmonary artery due to stenosis or a systemic-to-pulmonary arterial or cavopulmonary shunt. Pulmonary arteriovenous shunts and malformations (AVM). Pulmonary AVMs may be congenital or acquired. Hemodynamically significant AVMs should be embolized (31). Larger AVMs can be shown by MRA or CT angiography, while small AVMs are difficult to visualize even with

JACC Vol. 49, No. 9, 2007 March 6, 2007:993–1002

X-ray angiography. Magnetic resonance imaging experience with pulmonary AVMs and arteriovenous shunts is limited, but venous flow patterns are likely to be altered in the presence of intrapulmonary shunts. Magnetic resonance imaging-based determination of %HbO2 (21) in the pulmonary veins may, in principle, show reduced values in the presence of AVM, but has not yet been documented in the literature. Pulmonary veins and radiofrequency ablation. The pulmonary vein ostia are frequent sites of re-entry circuits in atrial fibrillation, and, therefore, present targets of radiofrequency ablation procedures, either by selective segmental ostial ablation or, increasingly, by circumferential pulmonary veins ablation (32,33). The 3D reconstructions from CT or MRI studies are used to guide proper positioning of the ablation catheter tip. Pulmonary vein imaging after ablation procedures has been advocated to rule out injury of the veins and to detect asymptomatic late stenosis (34). In our recent experience in a limited number of cases, the site and extent of ablation can be evaluated by using delayed enhancement, as has recently been demonstrated in dogs (35). Other clinical applications. Pulmonary vein or AVM thrombosis is a rare cause for cryptogenic stroke, especially in young patients, in whom other causes such as atrial fibrillation and atherosclerotic disease are rare (36). Highresolution pulmonary venography using multi-bolus, multiphase MRA has been used to rule out thrombi in the pulmonary veins (37). Mediastinal neoplasias can both infiltrate and externally obstruct the pulmonary veins (38). Aneurysms of the thoracic aorta may wedge the left pulmonary veins between the descending aorta and the heart (38). All of the above can mimic intrinsic pulmonary vein stenosis, resulting in a loss of phasic blood flow and high flow velocities, with the risk of pulmonary edema (39). Future Perspective

Figure 8

Results of a Bidirectional Cavopulmonary Anastomosis in a 14-Month-Old Boy With Right Isomerism

Contrast-enhanced angiograms show an unobstructed surgical anastomosis as well as the pulmonary arteries and veins. The right pulmonary artery (RPA) and vein are smaller than the vessels on the left side. The numbers in the figures represent flow volumes in liters per minute. There is net reversed flow in the left pulmonary artery (LPA). The amount of the systemic arterial collateral blood flow to each lung can be calculated by subtracting the pulmonary arterial blood flow volume from the pulmonary venous return. The right and left lungs receive systemic arterial collateral arterial flow of 0.2 and 0.6 l/min, respectively. Subsequent X-ray angiography confirmed florid systemic collateral circulation to the left lung and subtle collateral circulation to the right lung. SVC ⫽ superior vena cava. Reproduced with permission from Yoo et al. (30).

In the future, blood pool agents with a long intravascular half-life will further improve the spatial resolution of MRA and allow repeated image acquisitions under different patient postures, breathing maneuvers, and pharmacologic stimuli (40). Four-dimensional, or time-resolved MRA, will be used for the assessment of the changes in pulmonary perfusion secondary to pulmonary venous pathology as well as for the anatomical depiction of complex anatomy (41). Timeresolved velocity mapping that not only displays the anatomy in 3 dimensions but also traces blood flow in 3 directions simultaneously is now emerging and is comparable to the color Doppler mode of echocardiography (42). This technique outlines not only the origins of flow acceleration but also shows the streamlines of the turbulent blood flow encoded with different colors according to the velocity. Hyperpolarized noble gases such as 3-Helium and 129Xenon have been used for lung ventilation and gas exchange

JACC Vol. 49, No. 9, 2007 March 6, 2007:993–1002

(diffusion) (43). This technique may be helpful in the assessment of those pulmonary vein pathologies that cause ventilation perfusion mismatch, such as chronic pulmonary veno-occlusive disease, individual pulmonary vein stenosis, and pulmonary AVMs. Study limitations. The well-known general limitations of MRI such as a need for general anesthesia in small children, arrhythmias, and pacemakers apply to pulmonary vein imaging. In patients with very complex pulmonary venous anatomy, such as very short common trunks with early branching and tortuous veno-venous collaterals, exact flow quantification can be challenging, especially in neonates and small infants. Stenosis generates turbulence that causes dephasing artifacts in the PC MRI measurements. Furthermore, accurate flow quantification is usually impossible inside a stented vessel (44). One study in adults indicated that pulmonary vein ostia alternate in size by up to 24% throughout the cardiac cycle (45). This may be a combined effect of real changes in diameter and in vessel movement into and out of the imaging plane during systole and diastole. The authors also suggested that MRA, which is non-electrocardiogram gated, is subject to blurring and, therefore, tends to overestimate ostia size as compared with 2D cine imaging (45). Conclusions Contrast-enhanced MRA is now the gold standard for complete anatomic and functional diagnosis of anomalous pulmonary venous connections and pulmonary vein stenoses. In anomalous pulmonary venous connection, the amount of anomalous drainage and the Qp/Qs can be calculated by using PC MRI. Pulmonary venous flow velocity, volume, and patterns yield important information that is difficult or impossible to obtain using other modalities. Thus, MRA and PC MRI, in conjunction with cine imaging sequences, make MRI a powerful tool in the examination of the pulmonary circulation. Reprint requests and correspondence: Dr. Shi-Joon Yoo, Department of Diagnostic Imaging, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada M5G 1X8. E-mail: [email protected]. REFERENCES

1. 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;181:645–9. 2. Valsangiacomo E, Levasseur S, McCrindle B, MacDonald C, Smallhorn J, Yoo S. Contrast-enhanced MR angiography of pulmonary venous abnormalities in children. Pediatr Radiol 2003;33:92– 8. 3. Valsangiacomo E, Barrea C, Macgowan C, Smallhorn J, Coles J, Yoo S. Phase-contrast MR assessment of pulmonary venous blood flow in children with surgically repaired pulmonary veins. Pediatr Radiol 2003;33:607–13. 4. 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– 41.

Grosse-Wortmann et al. MRI Evaluation of Pulmonary Veins

1001

5. Fuchs F, Laub G, Othomo K. TrueFISP—technical considerations and cardiovascular applications. Eur J Radiol 2003;46:28 –32. 6. Hui BK, Noga ML, Gan KD, Wilman AH. Navigator-gated threedimensional MR angiography of the pulmonary arteries using steadystate free precession. J Magn Reson Imaging 2005;21:831–5. 7. Wu Y, Goodrich KC, Buswell HR, Katzman GL, Parker DL. High-resolution time-resolved contrast-enhanced 3D MRA by combining SENSE with keyhole and SLAM strategies. Magn Reson Imaging 2004;22:1161– 8. 8. Ohno Y, Hatabu H, Murase K, et al. Quantitative assessment of regional pulmonary perfusion in the entire lung using threedimensional ultrafast dynamic contrast-enhanced magnetic resonance imaging: preliminary experience in 40 subjects. J Magn Reson Imaging 2004;20:353– 65. 9. Ohno Y, Hatabu H, Murase K, et al. Quantitative assessment of regional pulmonary perfusion in the entire lung using threedimensional ultrafast dynamic contrast-enhanced magnetic resonance imaging: preliminary experience in 40 subjects. J Magn Reson Imaging 2004;20:353– 65. 10. Powell AJ, Maier SE, Chung T, Geva T. Phase-velocity cine magnetic resonance imaging measurement of pulsatile blood flow in children and young adults: in vitro and in vivo validation. Pediatr Cardiol 2000;21:104 –10. 11. Lotz J, Meier C, Leppert A, Galanski M. Cardiovascular flow measurement with phase-contrast MR imaging: basic facts and implementation. Radiographics 2002;22:651–71. 12. Beerbaum P, Korperich H, Gieseke J, Barth P, Peuster M, Meyer H. Blood flow quantification in adults by phase-contrast MRI combined with SENSE—a validation study. J Cardiovasc Magn Reson 2005;7: 361–9. 13. Kang IS, Redington AN, Benson LN, et al. Differential regurgitation in branch pulmonary arteries after repair of tetralogy of Fallot: a phase-contrast cine magnetic resonance study. Circulation 2003;107: 2938 – 43. 14. Hundley WG, Li HF, Hillis LD, et al. Quantitation of cardiac output with velocity-encoded, phase-difference magnetic resonance imaging. Am J Cardiol 1995;75:1250 –5. 15. Kellenberger CJ, Macgowan CK, Roman KS, et al. Hemodynamic evaluation of the peripheral pulmonary circulation by cine phasecontrast magnetic resonance imaging. J Magn Reson Imaging 2005; 22:780 –7. 16. Macgowan CK, Kellenberger CJ, Detsky JS, Roman K, Yoo SJ. Real-time Fourier velocity encoding: an in vivo evaluation. J Magn Reson Imaging 2005;21:297–304. 17. Klein AL, Leung DY, Murray RD, Urban LH, Bailey KR, Tajik AJ. Effects of age and physiologic variables on right ventricular filling dynamics in normal subjects. Am J Cardiol 1999;84:440 – 8. 18. Smiseth OA, Thompson CR, Lohavanichbutr K, et al. The pulmonary venous systolic flow pulse—its origin and relationship to left atrial pressure. J Am Coll Cardiol 1999;34:802–9. 19. Thulborn KR, Waterton JC, Matthews PM, Radda GK. Oxygenation dependence of the transverse relaxation time of water protons in whole blood at high field. Biochim Biophys Acta 1982;714:265–70. 20. Wright GA, Hu BS, Macovski A. 1991 I.I. Rabi Award. Estimating oxygen saturation of blood in vivo with MR imaging at 1.5 T. J Magn Reson Imaging 1991;1:275– 83. 21. Nield L, Qi X, Yoo S, Valsangiacomo E, Hornberger L, Wright G. MRI-based oxygen saturation measurements in infants and children with congenital heart disease. Pediatr Radiol 2002;32:518 –22. 22. Gilljam T, McCrindle BW, Smallhorn JF, Williams WG, Freedom RM. Outcomes of left atrial isomerism over a 28-year period at a single institution. J Am Coll Cardiol 2000;36:908 –16. 23. Festa P, Ait-Ali L, Cerillo AG, De Marchi MD, Murzi B. Magnetic resonance imaging is the diagnostic tool of choice in the preoperative evaluation of patients with partial anomalous pulmonary venous return. Int J Cardiovasc Imaging 2006;22:685–93. 24. Diethelm L, Dery R, Lipton MJ, Higgins CB. Atrial-level shunts: sensitivity and specificity of MR in diagnosis. Radiology 1987;162: 181– 6. 25. Khan MA, Torres AJ, Printz BF, Prakash A. Usefulness of magnetic resonance angiography for diagnosis of scimitar syndrome in early infancy. Am J Cardiol 2005;96:1313– 6.

1002

Grosse-Wortmann et al. MRI Evaluation of Pulmonary Veins

26. Devaney EJ, Chang AC, Ohye RG, Bove EL. Management of congenital and acquired pulmonary vein stenosis. Ann Thorac Surg 2006;81:992–5. 27. Roman KS, Kellenberger CJ, Macgowan CK, et al. How is pulmonary arterial blood flow affected by pulmonary venous obstruction in children? A phase-contrast magnetic resonance study. Pediatr Radiol 2005;35:580 – 6. 28. Bogren HG, Klipstein RH, Mohiaddin RH, et al. Pulmonary artery distensibility and blood flow patterns: a magnetic resonance study of normal subjects and of patients with pulmonary arterial hypertension. Am Heart J 1989;118:990 –9. 29. McElhinney DB, Reddy VM, Hanley FL, Moore P. Systemic venous collateral channels causing desaturation after bidirectional cavopulmonary anastomosis: evaluation and management. J Am Coll Cardiol 1997;30:817–24. 30. Yoo SJ, Kellenberger CJ, Roman KS, et al. Magnetic resonance evaluation of pulmonary circulation in children. Progr Pediatr Cardiol 2006;22:211–23. 31. Pollak JS, Saluja S, Thabet A, Henderson KJ, Denbow N, White RI Jr. Clinical and anatomic outcomes after embolotherapy of pulmonary arteriovenous malformations. J Vasc Interv Radiol 2006;17:35– 45. 32. Kato R, Lickfett L, Meininger G, et al. Pulmonary vein anatomy in patients undergoing catheter ablation of atrial fibrillation: lessons learned by use of magnetic resonance imaging. Circulation 2003;107:2004 –10. 33. Oral H, Pappone C, Chugh A, et al. Circumferential pulmonary-vein ablation for chronic atrial fibrillation. N Engl J Med 2006;354:934 – 41. 34. Tamborero D, Mont L, Nava S, et al. Incidence of pulmonary vein stenosis in patients submitted to atrial fibrillation ablation: a comparison of the selective segmental ostial ablation vs the circumferential pulmonary veins ablation. J Interv Card Electrophysiol 2005;14:21–5. 35. Dickfeld T, Kato R, Zviman M, et al. Characterization of radiofrequency ablation lesions with gadolinium-enhanced cardiovascular magnetic resonance imaging. J Am Coll Cardiol 2006;47:370 – 8.

JACC Vol. 49, No. 9, 2007 March 6, 2007:993–1002 36. Sloop RD, Lium JH. Systemic arterial embolism arising from pulmonary thrombophlebitis. Am J Surg 1971;37:503–5. 37. Grau AJ, Schoenberg SO, Lichy C, Buggle F, Bock M, Hacke W. Lack of evidence for pulmonary venous thrombosis in cryptogenic stroke: a magnetic resonance angiography study. Stroke 2002;33: 1416 –9. 38. Eicher JC, de Nadai L, Falcon-Eicher S, et al. [Unexplained pulmonary edema: demonstration of obstruction to pulmonary venous return by transesophageal echocardiography. Apropos of 4 cases.] Arch Mal Coeur Vaiss 1997;90:67–74. 39. Chen CL, Tunick PA, Kronzon I. Pulmonary vein compression by tumor: an unusual Doppler flow pattern. Echocardiography 2005;22: 746 –7. 40. Sirlin CB, Vera DR, Corbeil JA, Caballero MB, Buxton RB, Mattrey RF. Gadolinium-DTPA-dextran: a macromolecular MR blood pool contrast agent. Acad Radiol 2004;11:1361–9. 41. Zhu H, Buck DG, Zhang Z, et al. High temporal and spatial resolution 4D MRA using spiral data sampling and sliding window reconstruction. Magn Reson Med 2004;52:14 – 8. 42. Markl M, Draney MT, Miller DC, et al. Time-resolved threedimensional magnetic resonance velocity mapping of aortic flow in healthy volunteers and patients after valve-sparing aortic root replacement. J Thorac Cardiovasc Surg 2005;130:456 – 63. 43. van Beek EJ, Wild JM, Kauczor HU, Schreiber W, Mugler JP III, de Lange EE. Functional MRI of the lung using hyperpolarized 3-helium gas. J Magn Reson Imaging 2004;20:540 –54. 44. Wang Y, Truong TN, Yen C, et al. Quantitative evaluation of susceptibility and shielding effects of nitinol, platinum, cobalt-alloy, and stainless steel stents. Magn Reson Med 2003;49:972– 6. 45. Syed MA, Peters DC, Rashid H, Arai AE. Pulmonary vein imaging: comparison of 3D magnetic resonance angiography with 2D cine MRI for characterizing anatomy and size. J Cardiovasc Magn Reson 2005;7:355– 60.