- Email: [email protected]
Identification of efferent flow in the superior vena cava and azygos vein confluence using cine phase-contrast MRI: speculation of the role of the azygos arch valves
Available online at www.sciencedirect.com
Magnetic Resonance Imaging 28 (2010) 1306 – 1310
Identification of efferent flow in the superior vena cava and azygos vein confluence using cine phase-contrast MRI: speculation of the role of the azygos arch valves Satoru Morita⁎, Kazufumi Suzuki, Ai Masukawa, Shinya Kojima, Masami Hirata, Eiko Ueno Department of Radiology, Tokyo Women's Medical University Medical Center East, Tokyo 116-8567, Japan Received 25 November 2009; revised 23 January 2010; accepted 10 June 2010
Abstract Purpose: We aimed to evaluate flow patterns in the superior vena cava (SVC) and azygos vein confluence with cine phase-contrast magnetic resonance imaging with consideration for the role played by the azygos arch valves. Materials and Methods: Two-dimensional cine phase-contrast magnetic resonance images of the SVC and azygos vein confluence were prospectively acquired in 10 healthy volunteers. Flow directions during the cardiac cycle were evaluated quantitatively using sequential flow profile graphs obtained from each orthogonal image and affirmed visually by two radiologists from the oblique sagittal cine images. Results: Although the blood in the SVC and azygos vein confluence had an afferent flow during the systolic phase, a slight temporal efferent flow during the diastolic phase was quantitatively observed in all cases. Flow in the SVC can also be confirmed visually. The average velocity, average maximum afferent velocity during the systolic phase and average maximum efferent velocity during the diastolic phase of the SVC were 8.7±2.4, 19.9±3.7 and −1.0±3.2 cm/s, respectively; for the azygos vein confluence, these values were 2.2±1.5, 7.1±2.6 and −1.5±1.1 cm/s, respectively. Conclusion: We verified that a slight temporal efferent flow exists in the SVC and azygos vein confluence during the diastolic phase, which suggests that the usual role of the azygos arch valves is to prevent this physiological retrograde flow. © 2010 Elsevier Inc. All rights reserved. Keywords: Azygos arch valve; Azygos vein; MRI; Phase contrast; Blood flow
1. Introduction The azygos venous system consists of veins on each side of the vertebral column that drain the back and the thoracic and abdominal walls. It includes the azygos vein, hemiazygos vein and accessory hemiazygos vein. These veins join together and form the azygos arch, which finally drains into the superior vena cava (SVC) at approximately the level of the fourth and fifth thoracic vertebral bodies [1,2]. The azygos venous system plays an important role as a collateral pathway in the event of vena cava occlusion [1,2]. It is known that bicuspid valves exist in the azygos arch [3,4]. In patients with high central venous pressure, such as from cardiac tamponade, the azygos arch valves are
⁎ Corresponding author. Tel.: +81 3 3810 1111; fax: +81 3 3894 0282. E-mail address: [email protected] (S. Morita). 0730-725X/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.mri.2010.06.005
sometimes detected by computed tomography (CT) due to the reflux of contrast material into the azygos vein beyond the valves [5]. Even without such high central venous pressure, these valves are commonly observed, due to the development of multi-detector computed tomography (MDCT), especially following high rate injection of contrast material [3,4,6]. However, the reason why such reflux into the azygos vein occurs has not been fully explained and the role and significance of the azygos arch valves remain to be elucidated. Using cine phase-contrast magnetic resonance imaging (MRI), it is possible to noninvasively evaluate blood flow directions and velocities in various vessels in vivo [7–11]. Some reports describe the utility of this method in evaluating flow velocity in the azygos veins of patients with portal hypertension [7–11]. However, to date, there have been no studies to clarify the role of the azygos arch valves by evaluating flow patterns in the SVC and the azygos arch.
S. Morita et al. / Magnetic Resonance Imaging 28 (2010) 1306–1310
1307
We hypothesized that efferent blood flow can occur from the SVC to the azygos vein confluence and that the role of the azygos arch valves is to prevent this retrograde flow. Therefore, the purpose of the present study was to prospectively evaluate the blood flow patterns in the SVC and the azygos vein confluence using cine phase-contrast MRI, with a view to speculating on the role played by the azygos arch valves. 2. Materials and methods 2.1. Subjects The prospective single-institution study was approved by the Institutional Review Board of this facility. Informed written consent was obtained from each of the 10 healthy volunteers (7 men and 3 women; median age, 33 years; range, 25–58 years) with no symptoms or history (including data on the laboratory examination) of cardiovascular diseases or portal hypertension. 2.2. MRI technique MRI examinations were performed using a commercially available 1.5-T system (Magnetom Avanto, Siemens Medical Solutions, Erlangen, Germany). A six-element body phased-array coil, along with a six-element spine-array coil, was connected to a four-channel radiofrequency receiver. After obtaining two-dimensional trueFISP transaxial and sagittal localizing images, two-dimensional, retrospectively pulse-gated, spoiled gradient-echo-based cine phase-contrast sequences were acquired from subjects during breathholding after expiration. The oblique sagittal plane of the SVC, including the azygos vein confluence, oblique transaxial plane of the SVC perpendicular to it at the level of the azygos vein confluence and the oblique coronal plane of the azygos vein confluence perpendicular to it, was obtained (Fig. 1). Velocity encoding for the SVC and the azygos vein confluence were set to 80 and 10–15 cm/s, respectively, in all directions and adapted in the case of aliasing. Each set was reconstructed to yield magnitude images and velocity-encoded phase-contrast images in each of the three directions. The other acquisition parameters were as follows: time of repetition (TR)/time of echo (TE), 98.5/ 5.7 ms; flip angle, 15°; bandwidth, 698 Hz/pixel; matrix, 256×205 (interpolated to 512×410 matrix); field of view, 340×276 mm; section thickness, 5 mm; number of k-space segments (echo train length), 3; time resolution, 20 frames in one cardiac cycle; number of slices, 1; and acquisition time, approximately 47 R–R intervals. GRAPPA (generalized auto-calibrating partially parallel acquisition) with a parallel acquisition factor of 2 and 24 reference lines was used. 2.3. Image processing and analysis Blood flow velocities in the SVC and the azygos vein confluence during the cardiac cycle were quantitatively calculated using standard software on the console (Syn-
Fig. 1. Localizing two-dimensional trueFISP (A) transaxial and (B) sagittal images showing the cross sections of the oblique sagittal plane of the SVC (arrow) including the azygos vein confluence (arrowhead), oblique transaxial plane of the SVC at the azygos vein confluence level and the oblique coronal plane of the azygos vein confluence perpendicular to it for two-dimensional cine phase-contrast magnetic resonance images.
goMR B15, Siemens Medical Solutions). Flow velocities were obtained by locating a region of interest (ROI) on the orthogonal cine phase-contrast images of each vein reconstructed with a through-plane velocity-encoded direction. A radiologist (S.M) with more than 5 years' experience of evaluating MRI selected the ROI. Sequential flow profiles of
1308
S. Morita et al. / Magnetic Resonance Imaging 28 (2010) 1306–1310
the SVC and the azygos vein confluence during the cardiac cycle were displayed as graphs. Two radiologists (K.S and A.M), each with more than 5 years' experience of evaluating MRI, evaluated blood flow directions during the cardiac cycle by consensus. The average velocity, average maximum afferent velocity and average maximum efferent velocity, if observed, were calculated automatically with the software on the console. The two radiologists (K.S and A.M) visually evaluated the blood flow directions in the SVC and the azygos vein confluence by consensus using a commercially available viewer (INFINITT PACS; INFINITT, Seoul, Korea). The oblique sagittal cine phase-contrast images, reconstructed with a foot-
to-head velocity-encoded direction, were evaluated using the cine mode display. In these images, a black color indicated that the blood was flowing caudally, which implied that the flow direction in the SVC was afferent. In contrast, a white color indicated that the blood was flowing cranially, which implied that the flow direction in the SVC was efferent. If an efferent flow was observed, its cardiac phase was recorded.
3. Results The flow profile graphs indicated that the blood flow in the SVC was afferent during the systolic phase in all cases
Fig. 2. Examples of phase-contrast magnetic resonance images from a healthy 34-year-old male volunteer. (A) The flow profiles of the SVC and azygos vein confluence during the cardiac cycle indicate that temporal efferent flow exists during the diastolic phase. (B) The black color of the SVC (arrowhead) indicates that the flow direction in the SVC is afferent in the systolic-phase image, which is implied by the white color of the ascending aorta (arrow). (C) The faint white color of the SVC (arrowhead) indicates that the flow direction of the SVC is efferent in the diastolic-phase image. (D) Cine mode display indicates that an afferent blood flow in the SVC exists during the systolic phase and that a temporal mild efferent flow exists during the diastolic phase.
S. Morita et al. / Magnetic Resonance Imaging 28 (2010) 1306–1310 Table 1 Average flow velocities in the SVC and azygos vein confluence Average velocity Average maximum Average maximum afferent velocity efferent velocity during systole during diastole SVC 8.7±2.4 Azygos vein 2.2±1.5
19.9±3.7 7.1±2.6
−1.0±3.2 −1.5±1.1
Data are average velocities (expressed in centimeters per second).
(Fig. 2A). A mild temporal efferent flow was also indicated during the diastolic phase in all cases (Fig. 2A). This could be observed visually on the sagittal cine phase-contrast images (Figs. 2B–D). As for the azygos vein confluence, the flow profile graph indicated the same flow pattern as that observed in the SVC, with a faint temporal efferent flow during the diastolic phase in all cases (Fig. 2A). The efferent flow of blood in the azygos vein confluence could not be observed visually on the sagittal cine phase-contrast images (Figs. 2B–D). The average velocity, average maximum afferent velocity during the systolic phase and average maximum efferent velocity during the diastolic phase of the SVC and the azygos vein confluence are shown in Table 1.
4. Discussion Using cine phase-contrast MRI, we visually and quantitatively verified that a temporal mild physiological efferent blood flow exists in the SVC during the diastolic phase. This phenomenon is caused by atrial contraction, which pushes blood in the right atrium into the right ventricle during the diastolic phase. At the same time, some of the blood is pushed antidromically towards the SVC, which produces a retrograde flow in the SVC. The flow profile graph indicated a similar faint temporal efferent flow in the azygos vein confluence during the cardiac cycle. This would be preceded by an efferent flow of blood in the SVC. Its continuity could not be confirmed visually on the sagittal cine phase-contrast images. The reason for this would be due to the fact that the velocities of blood flow in the azygos vein confluence were too slow to be observed with the high velocity-encoding setting used to evaluate the SVC. The efferent flow from the SVC to the azygos vein confluence suggests that the usual role of the azygos arch valves is to prevent this physiological retrograde blood flow. If the azygos arch valves were not present, the direction of the blood flow in the azygos venous system might invert, because this efferent flow could potentially overcome the weaker afferent flow in the distal azygos vein. Thus, the azygos arch valves appear to play an important role in preserving the balance between the systemic venous system and the azygos system. Past reports have revealed that reflux of contrast material into the azygos arch is commonly seen in CT images,
1309
especially when the injection rate was high [3,4]. The findings of the present study indicate that the reason for this is that the contrast material clashes with the efferent blood flow in the SVC during the diastolic phase and refluxes into the azygos vein. If the injection rate is high, the degree of reflux becomes more pronounced as indicated in previous reports [3,4]. Then, the reflux of contrast material will generally be dammed by the azygos arch valves. The reflux of contrast material into the azygos vein beyond the azygos arch valves is often observed in patients with cardiac tamponade, as well as in patients with a variety of conditions that raise the central venous pressure [5]. In such cases, efferent flow in the SVC will increase due to the increase in pressure in the right side of the heart. When the blood pressure in the azygos vein confluence exceeds the holding power of the valves, the valves might open. This speculation cannot be proven because we did not evaluate the flow patterns in these patients. A past study has reported a high frequency of contrast material reflux posterior to the azygos arch valves (53.2%), which was recognized as a valve insufficiency [6]. This finding even raises the possibility that the azygos arch valve is not a true functional valve but a dysfunctional valve [6]. However, in common with the present study, this report could also not reach a conclusion as to whether this finding was caused by the valve insufficiency or the increase in pressure in the right side of the heart, due to the study's lack of patient population evaluation. Another limitation of the present study is that only a small number of healthy subjects were evaluated, and parameters such as respiration state and body position were not varied during the study. As for respiration, we obtained images during breath-holding after expiration. The flow patterns of the SVC and azygos vein may differ during free breathing or inspiration, because the flow velocity of the SVC and azygos vein generally increases during inspiration [9]. Similarly, these flow patterns will change depending on the body position, among others. An additional limitation of our study is that we did not directly visualize the movement of the azygos arch valves. Namely, our statement as to the role of the azygos valves is just speculation based on the indirect findings of flow pattern. However, we believe that this speculation is reasonable given the location of the azygos arch valves, which are within 4 cm of the SVC [3]. In addition, we did not visualize the three-dimensional flow dynamics around the azygos arch valves. In the future, it might be possible to evaluate the azygos arch valves in greater detail due to technical improvements in MRI. In conclusion, we confirmed that a slight temporal physiological efferent blood flow exists in the SVC and azygos vein confluence during the diastolic phase. This finding provides an explanation for the common observation of contrast material reflux into the azygos arch on MDCT. Furthermore, our results suggest that the usual role of the azygos arch valves is to prevent this
1310
S. Morita et al. / Magnetic Resonance Imaging 28 (2010) 1306–1310
physiological retrograde blood flow. Thus, these valves may play an important role in preserving and adjusting the balance between the systemic venous system and the azygos system.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.mri .2010.06.005.
References [1] Smathers RL, Buschi AJ, Pope Jr TL, Brenbridge AN, Williamson BR. The azygous arch: normal and pathologic CT appearance. AJR Am J Roentgenol 1982;139:477–83. [2] Dudiak CM, Olson MC, Posniak HV. CT evaluation of congenital and acquired abnormalities of the azygos system. Radiographics 1991;11:233–46. [3] Yeh BM, Coakley FV, Sanchez HC, Wilson MW, Reddy GP, Gotway MB. Azygos arch valves: prevalence and appearance at contrastenhanced CT. Radiology 2004;230:111–5.
[4] Ichikawa T, Endo J, Koizumi J, Ro A, Kobayashi M, Saito M, et al. Visualization of the azygos arch valves on multidetector-row computed tomography. Heart Vessels 2008;23:118–23. [5] Restrepo CS, Lemos DF, Lemos JA, Velasquez E, Diethelm L, Ovella TA, et al. Imaging findings in cardiac tamponade with emphasis on CT. Radiographics 2007;27:1595–610. [6] Steinke K, Moghaddam A. Azygos arch valves at computed tomography angiography and pitfalls related to its variety in appearance and function. J Comput Assist Tomogr 2009;33:721–4. [7] Wu MT, Pan HB, Chen C, Chang JM, Lo GH, Wu SS, et al. Azygos blood flow in cirrhosis: measurement with MR imaging and correlation with variceal hemorrhage. Radiology 1996;198:457–62. [8] Lomas DJ, Hayball MP, Jones DP, Sims C, Allison ME, Alexander GJ. Non-invasive measurement of azygos venous blood flow using magnetic resonance. J Hepatol 1995;22:399–403. [9] Ng WH, Chan YL, Sung JY, Lee YT, Lee SF, Chung SS. Comparison of breath-hold 2D phase-contrast with non breath-hold cine phasecontrast MRA in the assessment of azygos venous blood flow in portal hypertension. MAGMA 2004;16:211–7. [10] Sugano S, Yamamoto K, Takamura N, Momiyama K, Watanabe M, Ishii K. Azygos venous blood flow while fasting, postprandially, and after endoscopic variceal ligation, measured by magnetic resonance imaging. J Gastroenterol 1999;34:310–4. [11] Debatin JF, Zahner B, Meyenberger C, Romanowski B, Schöpke W, Marincek B, et al. Azygos blood flow: phase contrast quantitation in volunteers and patients with portal hypertension pre- and postintrahepatic shunt placement. Hepatology 1996;24:1109–15.
Magnetic Resonance Imaging 28 (2010) 1306 – 1310
Identification of efferent flow in the superior vena cava and azygos vein confluence using cine phase-contrast MRI: speculation of the role of the azygos arch valves Satoru Morita⁎, Kazufumi Suzuki, Ai Masukawa, Shinya Kojima, Masami Hirata, Eiko Ueno Department of Radiology, Tokyo Women's Medical University Medical Center East, Tokyo 116-8567, Japan Received 25 November 2009; revised 23 January 2010; accepted 10 June 2010
Abstract Purpose: We aimed to evaluate flow patterns in the superior vena cava (SVC) and azygos vein confluence with cine phase-contrast magnetic resonance imaging with consideration for the role played by the azygos arch valves. Materials and Methods: Two-dimensional cine phase-contrast magnetic resonance images of the SVC and azygos vein confluence were prospectively acquired in 10 healthy volunteers. Flow directions during the cardiac cycle were evaluated quantitatively using sequential flow profile graphs obtained from each orthogonal image and affirmed visually by two radiologists from the oblique sagittal cine images. Results: Although the blood in the SVC and azygos vein confluence had an afferent flow during the systolic phase, a slight temporal efferent flow during the diastolic phase was quantitatively observed in all cases. Flow in the SVC can also be confirmed visually. The average velocity, average maximum afferent velocity during the systolic phase and average maximum efferent velocity during the diastolic phase of the SVC were 8.7±2.4, 19.9±3.7 and −1.0±3.2 cm/s, respectively; for the azygos vein confluence, these values were 2.2±1.5, 7.1±2.6 and −1.5±1.1 cm/s, respectively. Conclusion: We verified that a slight temporal efferent flow exists in the SVC and azygos vein confluence during the diastolic phase, which suggests that the usual role of the azygos arch valves is to prevent this physiological retrograde flow. © 2010 Elsevier Inc. All rights reserved. Keywords: Azygos arch valve; Azygos vein; MRI; Phase contrast; Blood flow
1. Introduction The azygos venous system consists of veins on each side of the vertebral column that drain the back and the thoracic and abdominal walls. It includes the azygos vein, hemiazygos vein and accessory hemiazygos vein. These veins join together and form the azygos arch, which finally drains into the superior vena cava (SVC) at approximately the level of the fourth and fifth thoracic vertebral bodies [1,2]. The azygos venous system plays an important role as a collateral pathway in the event of vena cava occlusion [1,2]. It is known that bicuspid valves exist in the azygos arch [3,4]. In patients with high central venous pressure, such as from cardiac tamponade, the azygos arch valves are
⁎ Corresponding author. Tel.: +81 3 3810 1111; fax: +81 3 3894 0282. E-mail address: [email protected] (S. Morita). 0730-725X/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.mri.2010.06.005
sometimes detected by computed tomography (CT) due to the reflux of contrast material into the azygos vein beyond the valves [5]. Even without such high central venous pressure, these valves are commonly observed, due to the development of multi-detector computed tomography (MDCT), especially following high rate injection of contrast material [3,4,6]. However, the reason why such reflux into the azygos vein occurs has not been fully explained and the role and significance of the azygos arch valves remain to be elucidated. Using cine phase-contrast magnetic resonance imaging (MRI), it is possible to noninvasively evaluate blood flow directions and velocities in various vessels in vivo [7–11]. Some reports describe the utility of this method in evaluating flow velocity in the azygos veins of patients with portal hypertension [7–11]. However, to date, there have been no studies to clarify the role of the azygos arch valves by evaluating flow patterns in the SVC and the azygos arch.
S. Morita et al. / Magnetic Resonance Imaging 28 (2010) 1306–1310
1307
We hypothesized that efferent blood flow can occur from the SVC to the azygos vein confluence and that the role of the azygos arch valves is to prevent this retrograde flow. Therefore, the purpose of the present study was to prospectively evaluate the blood flow patterns in the SVC and the azygos vein confluence using cine phase-contrast MRI, with a view to speculating on the role played by the azygos arch valves. 2. Materials and methods 2.1. Subjects The prospective single-institution study was approved by the Institutional Review Board of this facility. Informed written consent was obtained from each of the 10 healthy volunteers (7 men and 3 women; median age, 33 years; range, 25–58 years) with no symptoms or history (including data on the laboratory examination) of cardiovascular diseases or portal hypertension. 2.2. MRI technique MRI examinations were performed using a commercially available 1.5-T system (Magnetom Avanto, Siemens Medical Solutions, Erlangen, Germany). A six-element body phased-array coil, along with a six-element spine-array coil, was connected to a four-channel radiofrequency receiver. After obtaining two-dimensional trueFISP transaxial and sagittal localizing images, two-dimensional, retrospectively pulse-gated, spoiled gradient-echo-based cine phase-contrast sequences were acquired from subjects during breathholding after expiration. The oblique sagittal plane of the SVC, including the azygos vein confluence, oblique transaxial plane of the SVC perpendicular to it at the level of the azygos vein confluence and the oblique coronal plane of the azygos vein confluence perpendicular to it, was obtained (Fig. 1). Velocity encoding for the SVC and the azygos vein confluence were set to 80 and 10–15 cm/s, respectively, in all directions and adapted in the case of aliasing. Each set was reconstructed to yield magnitude images and velocity-encoded phase-contrast images in each of the three directions. The other acquisition parameters were as follows: time of repetition (TR)/time of echo (TE), 98.5/ 5.7 ms; flip angle, 15°; bandwidth, 698 Hz/pixel; matrix, 256×205 (interpolated to 512×410 matrix); field of view, 340×276 mm; section thickness, 5 mm; number of k-space segments (echo train length), 3; time resolution, 20 frames in one cardiac cycle; number of slices, 1; and acquisition time, approximately 47 R–R intervals. GRAPPA (generalized auto-calibrating partially parallel acquisition) with a parallel acquisition factor of 2 and 24 reference lines was used. 2.3. Image processing and analysis Blood flow velocities in the SVC and the azygos vein confluence during the cardiac cycle were quantitatively calculated using standard software on the console (Syn-
Fig. 1. Localizing two-dimensional trueFISP (A) transaxial and (B) sagittal images showing the cross sections of the oblique sagittal plane of the SVC (arrow) including the azygos vein confluence (arrowhead), oblique transaxial plane of the SVC at the azygos vein confluence level and the oblique coronal plane of the azygos vein confluence perpendicular to it for two-dimensional cine phase-contrast magnetic resonance images.
goMR B15, Siemens Medical Solutions). Flow velocities were obtained by locating a region of interest (ROI) on the orthogonal cine phase-contrast images of each vein reconstructed with a through-plane velocity-encoded direction. A radiologist (S.M) with more than 5 years' experience of evaluating MRI selected the ROI. Sequential flow profiles of
1308
S. Morita et al. / Magnetic Resonance Imaging 28 (2010) 1306–1310
the SVC and the azygos vein confluence during the cardiac cycle were displayed as graphs. Two radiologists (K.S and A.M), each with more than 5 years' experience of evaluating MRI, evaluated blood flow directions during the cardiac cycle by consensus. The average velocity, average maximum afferent velocity and average maximum efferent velocity, if observed, were calculated automatically with the software on the console. The two radiologists (K.S and A.M) visually evaluated the blood flow directions in the SVC and the azygos vein confluence by consensus using a commercially available viewer (INFINITT PACS; INFINITT, Seoul, Korea). The oblique sagittal cine phase-contrast images, reconstructed with a foot-
to-head velocity-encoded direction, were evaluated using the cine mode display. In these images, a black color indicated that the blood was flowing caudally, which implied that the flow direction in the SVC was afferent. In contrast, a white color indicated that the blood was flowing cranially, which implied that the flow direction in the SVC was efferent. If an efferent flow was observed, its cardiac phase was recorded.
3. Results The flow profile graphs indicated that the blood flow in the SVC was afferent during the systolic phase in all cases
Fig. 2. Examples of phase-contrast magnetic resonance images from a healthy 34-year-old male volunteer. (A) The flow profiles of the SVC and azygos vein confluence during the cardiac cycle indicate that temporal efferent flow exists during the diastolic phase. (B) The black color of the SVC (arrowhead) indicates that the flow direction in the SVC is afferent in the systolic-phase image, which is implied by the white color of the ascending aorta (arrow). (C) The faint white color of the SVC (arrowhead) indicates that the flow direction of the SVC is efferent in the diastolic-phase image. (D) Cine mode display indicates that an afferent blood flow in the SVC exists during the systolic phase and that a temporal mild efferent flow exists during the diastolic phase.
S. Morita et al. / Magnetic Resonance Imaging 28 (2010) 1306–1310 Table 1 Average flow velocities in the SVC and azygos vein confluence Average velocity Average maximum Average maximum afferent velocity efferent velocity during systole during diastole SVC 8.7±2.4 Azygos vein 2.2±1.5
19.9±3.7 7.1±2.6
−1.0±3.2 −1.5±1.1
Data are average velocities (expressed in centimeters per second).
(Fig. 2A). A mild temporal efferent flow was also indicated during the diastolic phase in all cases (Fig. 2A). This could be observed visually on the sagittal cine phase-contrast images (Figs. 2B–D). As for the azygos vein confluence, the flow profile graph indicated the same flow pattern as that observed in the SVC, with a faint temporal efferent flow during the diastolic phase in all cases (Fig. 2A). The efferent flow of blood in the azygos vein confluence could not be observed visually on the sagittal cine phase-contrast images (Figs. 2B–D). The average velocity, average maximum afferent velocity during the systolic phase and average maximum efferent velocity during the diastolic phase of the SVC and the azygos vein confluence are shown in Table 1.
4. Discussion Using cine phase-contrast MRI, we visually and quantitatively verified that a temporal mild physiological efferent blood flow exists in the SVC during the diastolic phase. This phenomenon is caused by atrial contraction, which pushes blood in the right atrium into the right ventricle during the diastolic phase. At the same time, some of the blood is pushed antidromically towards the SVC, which produces a retrograde flow in the SVC. The flow profile graph indicated a similar faint temporal efferent flow in the azygos vein confluence during the cardiac cycle. This would be preceded by an efferent flow of blood in the SVC. Its continuity could not be confirmed visually on the sagittal cine phase-contrast images. The reason for this would be due to the fact that the velocities of blood flow in the azygos vein confluence were too slow to be observed with the high velocity-encoding setting used to evaluate the SVC. The efferent flow from the SVC to the azygos vein confluence suggests that the usual role of the azygos arch valves is to prevent this physiological retrograde blood flow. If the azygos arch valves were not present, the direction of the blood flow in the azygos venous system might invert, because this efferent flow could potentially overcome the weaker afferent flow in the distal azygos vein. Thus, the azygos arch valves appear to play an important role in preserving the balance between the systemic venous system and the azygos system. Past reports have revealed that reflux of contrast material into the azygos arch is commonly seen in CT images,
1309
especially when the injection rate was high [3,4]. The findings of the present study indicate that the reason for this is that the contrast material clashes with the efferent blood flow in the SVC during the diastolic phase and refluxes into the azygos vein. If the injection rate is high, the degree of reflux becomes more pronounced as indicated in previous reports [3,4]. Then, the reflux of contrast material will generally be dammed by the azygos arch valves. The reflux of contrast material into the azygos vein beyond the azygos arch valves is often observed in patients with cardiac tamponade, as well as in patients with a variety of conditions that raise the central venous pressure [5]. In such cases, efferent flow in the SVC will increase due to the increase in pressure in the right side of the heart. When the blood pressure in the azygos vein confluence exceeds the holding power of the valves, the valves might open. This speculation cannot be proven because we did not evaluate the flow patterns in these patients. A past study has reported a high frequency of contrast material reflux posterior to the azygos arch valves (53.2%), which was recognized as a valve insufficiency [6]. This finding even raises the possibility that the azygos arch valve is not a true functional valve but a dysfunctional valve [6]. However, in common with the present study, this report could also not reach a conclusion as to whether this finding was caused by the valve insufficiency or the increase in pressure in the right side of the heart, due to the study's lack of patient population evaluation. Another limitation of the present study is that only a small number of healthy subjects were evaluated, and parameters such as respiration state and body position were not varied during the study. As for respiration, we obtained images during breath-holding after expiration. The flow patterns of the SVC and azygos vein may differ during free breathing or inspiration, because the flow velocity of the SVC and azygos vein generally increases during inspiration [9]. Similarly, these flow patterns will change depending on the body position, among others. An additional limitation of our study is that we did not directly visualize the movement of the azygos arch valves. Namely, our statement as to the role of the azygos valves is just speculation based on the indirect findings of flow pattern. However, we believe that this speculation is reasonable given the location of the azygos arch valves, which are within 4 cm of the SVC [3]. In addition, we did not visualize the three-dimensional flow dynamics around the azygos arch valves. In the future, it might be possible to evaluate the azygos arch valves in greater detail due to technical improvements in MRI. In conclusion, we confirmed that a slight temporal physiological efferent blood flow exists in the SVC and azygos vein confluence during the diastolic phase. This finding provides an explanation for the common observation of contrast material reflux into the azygos arch on MDCT. Furthermore, our results suggest that the usual role of the azygos arch valves is to prevent this
1310
S. Morita et al. / Magnetic Resonance Imaging 28 (2010) 1306–1310
physiological retrograde blood flow. Thus, these valves may play an important role in preserving and adjusting the balance between the systemic venous system and the azygos system.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.mri .2010.06.005.
References [1] Smathers RL, Buschi AJ, Pope Jr TL, Brenbridge AN, Williamson BR. The azygous arch: normal and pathologic CT appearance. AJR Am J Roentgenol 1982;139:477–83. [2] Dudiak CM, Olson MC, Posniak HV. CT evaluation of congenital and acquired abnormalities of the azygos system. Radiographics 1991;11:233–46. [3] Yeh BM, Coakley FV, Sanchez HC, Wilson MW, Reddy GP, Gotway MB. Azygos arch valves: prevalence and appearance at contrastenhanced CT. Radiology 2004;230:111–5.
[4] Ichikawa T, Endo J, Koizumi J, Ro A, Kobayashi M, Saito M, et al. Visualization of the azygos arch valves on multidetector-row computed tomography. Heart Vessels 2008;23:118–23. [5] Restrepo CS, Lemos DF, Lemos JA, Velasquez E, Diethelm L, Ovella TA, et al. Imaging findings in cardiac tamponade with emphasis on CT. Radiographics 2007;27:1595–610. [6] Steinke K, Moghaddam A. Azygos arch valves at computed tomography angiography and pitfalls related to its variety in appearance and function. J Comput Assist Tomogr 2009;33:721–4. [7] Wu MT, Pan HB, Chen C, Chang JM, Lo GH, Wu SS, et al. Azygos blood flow in cirrhosis: measurement with MR imaging and correlation with variceal hemorrhage. Radiology 1996;198:457–62. [8] Lomas DJ, Hayball MP, Jones DP, Sims C, Allison ME, Alexander GJ. Non-invasive measurement of azygos venous blood flow using magnetic resonance. J Hepatol 1995;22:399–403. [9] Ng WH, Chan YL, Sung JY, Lee YT, Lee SF, Chung SS. Comparison of breath-hold 2D phase-contrast with non breath-hold cine phasecontrast MRA in the assessment of azygos venous blood flow in portal hypertension. MAGMA 2004;16:211–7. [10] Sugano S, Yamamoto K, Takamura N, Momiyama K, Watanabe M, Ishii K. Azygos venous blood flow while fasting, postprandially, and after endoscopic variceal ligation, measured by magnetic resonance imaging. J Gastroenterol 1999;34:310–4. [11] Debatin JF, Zahner B, Meyenberger C, Romanowski B, Schöpke W, Marincek B, et al. Azygos blood flow: phase contrast quantitation in volunteers and patients with portal hypertension pre- and postintrahepatic shunt placement. Hepatology 1996;24:1109–15.