Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print March 18, 2008, 10.1148/radiol.2472070790
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2472070790v1
247/2/558    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kim, C. Y.
Right arrow Articles by Merkle, E. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kim, C. Y.
Right arrow Articles by Merkle, E. M.
(Radiology 2008;247:558-566.)
© RSNA, 2008

Thoracic Imaging

Central Veins of the Chest: Evaluation with Time-resolved MR Angiography1

Charles Y. Kim, MD, Rizvan A. Mirza, MD, Joshua A. Bryant, MD, Eric D. Whiting, MD, David M. Delong, PhD, Charles E. Spritzer, MD, and Elmar M. Merkle, MD

1 From the Department of Radiology, Duke University Medical Center, 2301 Erwin Rd, Box 3808, Durham, NC 27710. Received May 6, 2007; revision requested July 6; revision received August 16; accepted September 12; final version accepted October 15. Address correspondence to E.M.M. (e-mail: elmar.merkle{at}duke.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
Purpose: To retrospectively assess the diagnostic performance of time-resolved magnetic resonance (MR) angiography in the detection of stenoses and occlusions in the central veins of the chest, with angiographic and surgical findings and consensus readings serving as the reference standard.

Materials and Methods: Institutional review board approval was obtained, and the informed consent requirement was waived for this HIPAA-compliant study. Retrospective analysis was performed with 27 consecutive patients (12 male, 15 female; age range, 16–67 years) who underwent MR venography of the central veins. Six radiologists with varying levels of experience interpreted the studies. For each study, the readers were presented with time-resolved maximum intensity projection (MIP) images only, high-spatial-resolution images only, or both. Sensitivity and specificity were calculated for detection of stenoses and occlusions, as well as for confidence levels, study interpretation time, and determination of the side of the body on which upper extremity contrast material injection was performed.

Results: The addition of time-resolved angiographic images to the high-spatial-resolution images resulted in improved specificity in the detection of venous occlusions (0.99 vs 0.96, P = .03), in reader confidence (P < .001), and in the ability to infer the side of injection (83% correct compared with 32% correct, P < .001), without increasing the average time required for study interpretation. Use of time-resolved angiographic data sets as a stand-alone technique had high sensitivity (0.95) but only moderate specificity (0.56) in the detection of venous stenoses or occlusions.

Conclusion: Time-resolved angiographic images are a useful adjunct to high-spatial-resolution images in the evaluation of central venous stenoses and occlusions.

© RSNA, 2008


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
Magnetic resonance (MR) imaging has become increasingly popular as a noninvasive means with which to examine the central venous system for stenoses and occlusions, and it is particularly useful in the examination of patients who may have superior vena cava syndrome and in access planning for central venous catheters and arteriovenous dialysis grafts (1). When compared with conventional angiography, contrast material–enhanced MR evaluation of the central veins has been reported to be extremely sensitive and specific in the detection of stenoses and occlusions (2–6). However, a large amount of gadolinium-based contrast material is required, and the dose typically exceeds the standard single dose used for most contrast-enhanced MR studies. In light of recent reports (7–11) that have implicated gadolinium-based contrast agents as a causative factor in the development of nephrogenic systemic fibrosis after high-dose MR examinations in patients with impaired renal function, it seems prudent to minimize the amount of gadolinium administered during central venous system assessment, particularly in this at-risk patient population.

Unlike conventionally acquired static high-spatial-resolution contrast-enhanced MR angiographic data sets, time-resolved MR images are acquired rapidly at short time intervals, thereby enabling visualization of the temporal dynamics of blood flow (12,13). Additionally, time-resolved data sets usually require substantially less contrast material than do high-spatial-resolution contrast-enhanced MR angiographic data sets. Thus, the purpose of this study was to retrospectively assess the diagnostic performance of the MR time-resolved echo-shared angiographic technique in the detection of stenoses and occlusions in the central veins of the chest, with conventional angiographic and surgical findings and consensus readings serving as the reference standard.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
Our study was not directly supported by industry; however, our institution has a research collaboration with Siemens Medical Solutions, Malvern, Pa. The authors had full control of the data and information submitted for publication.

Patients
From January through August 2006, a total of 27 consecutive patients (12 male, 15 female; age range, 16–67 years) underwent gadolinium-enhanced MR imaging of the chest to evaluate the central veins. At our institution, MR angiography is the first-line imaging modality used to evaluate superior vena cava syndrome and potential hemodialysis access sites in patients known to have substantial venous disease. All 27 patients were included in this retrospective study. The clinical indications included evaluation of upper extremity or facial swelling (n = 14), venous access planning (n = 9), and evaluation of possible vasculitis or venous anatomy (n = 4). Twelve patients had renal insufficiency or end-stage renal disease. Institutional review board approval and waiver of the informed consent requirement were obtained for this retrospective Health Insurance Portability and Accountability Act–compliant study.

MR Sequence Protocol
Imaging was performed with a magnetic field strength of 1.5 T (n = 10, Magnetom Symphony or Avanto; Siemens Medical Solutions, Erlangen, Germany) or 3 T (n = 17, Magnetom Tim Trio; Siemens Medical Solutions). Dedicated receive-only body-array coils (Siemens Medical Solutions) were used for signal reception. Patients were placed in the supine position with their arms at their sides. A large field of view (38–40 x 38–40 cm) and oversampling were used to avoid wraparound artifacts. After a three-plane localizer image was acquired, the time-resolved echo-shared angiographic sequence was applied in the coronal orientation (14–18) (Table 1). This sequence commenced 5 seconds after initiation of the first bolus administration of 10 mL of nondiluted gadopentetate dimeglumine (0.5 mmol/mL, Magnevist; Berlex, Wayne, NJ) at a rate of 2 mL/sec followed by administration of 20 mL of saline at a rate of 2 mL/sec. Image acquisition was repeated multiple times without delay for 60 seconds during shallow breathing. Contrast material was administered through a peripheral or central venous catheter, and the injection site was recorded on the nursing sheet. In patients with unilateral upper extremity swelling, the peripheral intravenous catheter was preferentially placed contralaterally. In patients in whom central venous catheters were used, a nurse manually injected the gadopentetate dimeglumine and saline bolus. Three-dimensional imaging data sets were processed into one coronal maximum intensity projection (MIP) image for each time point and stored as a cine loop consisting of up to 15 coronal images.


View this table:
[in this window]
[in a new window]

  		Table 1. MR Sequence Parameters for Time-resolved Echo-shared Angiographic Technique and High-Spatial-Resolution Contrast-enhanced Images

 
High-spatial-resolution contrast-enhanced images were obtained after intravenous administration of 30 mL of nondiluted gadopentetate dimeglumine at a rate of 2 mL/sec followed by 20 mL of saline, also administered at a rate of 2 mL/sec. A three-dimensional gradient-echo (turbo fast low-angle shot) sequence was applied in the coronal plane prior to administration of the second contrast material bolus. This sequence was repeated three more times after administration of the second contrast material bolus during a breath hold (Table 1). Data acquisition for the first high-spatial-resolution contrast-enhanced series was initiated 15–20 seconds after administration of the second contrast material bolus, with the exact delay time determined for the time-resolved imaging data set. Postcontrast data acquisition was repeated twice with a 15–20-second interval gap. The length of the interval gap depended on the patient's breath-holding capacity.

Image Analysis
Six radiologists with varying levels of experience interpreted the studies. The two senior-level faculty readers (E.M.M., C.E.S.) had 10 and 20 years of experience, respectively, in postsubspecialty training. Of the four junior readers, two were board-certified fellows in abdominal imaging (R.A.M., E.D.W.) and two were 2nd-year radiology residents (C.Y.K., J.A.B.).

Three separate reading sessions were conducted: In session 1, only time-resolved MIP images were read. In session 2, only high-spatial-resolution source images were read. In session 3, both time-resolved MIP images and high-spatial-resolution source images were read. The MR data sets were anonymized, and the specific images pertinent to each reading session were transferred to a freestanding workstation (Advantage Windows; GE Healthcare, Milwaukee, Wis). The studies were randomized only once prior to the initial reading session. The readers were blinded to all clinical and demographic information. The reading sessions were scheduled to be conducted at least 2 weeks apart to reduce recall bias. Each radiologist independently interpreted all studies.

For each session, one standardized evaluation form was supplied to each reader for each study to be interpreted. The start and finish times were rounded to the current minute and entered for each study. The reader recorded the side of the body on which the upper extremity injection was performed or if a central venous catheter was inserted; if this was unclear, the reader recorded an indeterminate finding. The central venous system of the chest was divided into seven venous segments comprising the internal jugular veins, subclavian veins, brachiocephalic veins, and superior vena cava; a total of 189 venous segments were included. The boundary between the subclavian vein and the axillary vein was defined 5 cm from the junction with the internal jugular vein. Each venous segment was evaluated for the highest degree of stenosis within that segment. Stenoses were graded as none to mild (<50% narrowing), moderate to severe (50%–95% narrowing), or subtotal to complete occlusion (>95% narrowing). Segments not depicted despite 60 seconds of circulation after injection were graded as occluded. The level of confidence for each venous segment assessment was graded as follows: 1, not confident (<50% certainty); 2, somewhat confident (50%–74% certainty); 3, moderately confident (75%–94% certainty); or 4, very confident (95%–100% certainty). The reference standard for the degree of stenosis in each segment was the interpretation of available conventional angiographic studies and notes from surgical procedures performed within 2 weeks of the MR examination. If none were available, the reference standard was determined with consensus reading performed by the two senior radiologists.

Statistical Analyses
A dedicated statistician (D.M.D.) designed and performed the statistical analyses. Analysis of study interpretation times was performed with a repeated-measures analysis of variance model. The side of injection responses were analyzed by comparing the average percentages correctly determined between sessions with use of the signed rank test. Reader-averaged sensitivity and specificity values for stenosis, occlusion, or both, were also compared by using the signed rank test. The interobserver agreement for rating stenoses or occlusions was measured with the linear-weighted Cohen {kappa} statistic for the two faculty readers and with the Fleiss {kappa} statistic for all readers. The confidence levels were analyzed by comparing average ratings over the venous segments between sessions with use of a signed rank test. Subset analyses were also performed for each of the three groups of readers. All data analyses were performed with SAS software (version 9.1.3; SAS Institute, Cary, NC).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
Of the 189 venous segments imaged, 12 had moderate to severe stenosis and 40 had subtotal to complete occlusion (Figs 1–4). Four patients had no significant stenosis or occlusion. Additional conventional angiographic imaging studies were available for nine patients, for a total of 36 venous segments; these studies consisted of eight venograms and one fistulogram. A surgical note was available for one central venous catheter placement, which provided information about three venous segments. Consensus readings were the reference standard for the remaining 150 venous segments.


Figure 1
View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Flowchart shows study inclusion and reference standards used. MRA = MR angiography.

 

Figure 2A
View larger version (128K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: Time-resolved coronal MIPs show subtle collateral vessels in a patient with moderate to severe stenosis of the right internal jugular vein. (a) Image acquired at an early time point shows gadolinium-based contrast agent inflow via the right cephalic (thick arrow) and subclavian (thin arrow) veins, with opacification of the right side of the heart and pulmonary arteries. (b) Image acquired during the arterial phase of contrast material administration shows predominantly arterial opacification. The right subclavian vein (arrow) remains opacified. (c) Image acquired during the venous phase of contrast material administration shows predominantly venous opacification. The left internal jugular vein (arrow) is of normal caliber, while the right internal jugular vein (*) is markedly stenotic. Asymmetric collateral flow through a mildly enlarged right external jugular vein (arrowhead) is consistent with a hemodynamically significant stenosis.

 

Figure 2B
View larger version (148K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: Time-resolved coronal MIPs show subtle collateral vessels in a patient with moderate to severe stenosis of the right internal jugular vein. (a) Image acquired at an early time point shows gadolinium-based contrast agent inflow via the right cephalic (thick arrow) and subclavian (thin arrow) veins, with opacification of the right side of the heart and pulmonary arteries. (b) Image acquired during the arterial phase of contrast material administration shows predominantly arterial opacification. The right subclavian vein (arrow) remains opacified. (c) Image acquired during the venous phase of contrast material administration shows predominantly venous opacification. The left internal jugular vein (arrow) is of normal caliber, while the right internal jugular vein (*) is markedly stenotic. Asymmetric collateral flow through a mildly enlarged right external jugular vein (arrowhead) is consistent with a hemodynamically significant stenosis.

 

Figure 2C
View larger version (141K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c: Time-resolved coronal MIPs show subtle collateral vessels in a patient with moderate to severe stenosis of the right internal jugular vein. (a) Image acquired at an early time point shows gadolinium-based contrast agent inflow via the right cephalic (thick arrow) and subclavian (thin arrow) veins, with opacification of the right side of the heart and pulmonary arteries. (b) Image acquired during the arterial phase of contrast material administration shows predominantly arterial opacification. The right subclavian vein (arrow) remains opacified. (c) Image acquired during the venous phase of contrast material administration shows predominantly venous opacification. The left internal jugular vein (arrow) is of normal caliber, while the right internal jugular vein (*) is markedly stenotic. Asymmetric collateral flow through a mildly enlarged right external jugular vein (arrowhead) is consistent with a hemodynamically significant stenosis.

 

Figure 3A
View larger version (116K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a: Time-resolved coronal MIPs show marked venous collateral vessels in a patient with a left internal jugular vein occlusion. (a) Image acquired during the early arterial phase of contrast material administration shows predominantly pulmonary and systemic arterial opacification. Some residual contrast material from the injection can be seen in the right subclavian vein. (b) Image acquired in the early venous phase shows normal opacification of the right internal jugular, subclavian, and brachiocephalic veins but with abrupt tapering of the caliber of the left internal jugular vein (*). (c) Marked collateralization in the region of the central left internal jugular vein by at least four venous collateral vessels (arrows) is consistent with a venous occlusion.

 

Figure 3B
View larger version (100K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b: Time-resolved coronal MIPs show marked venous collateral vessels in a patient with a left internal jugular vein occlusion. (a) Image acquired during the early arterial phase of contrast material administration shows predominantly pulmonary and systemic arterial opacification. Some residual contrast material from the injection can be seen in the right subclavian vein. (b) Image acquired in the early venous phase shows normal opacification of the right internal jugular, subclavian, and brachiocephalic veins but with abrupt tapering of the caliber of the left internal jugular vein (*). (c) Marked collateralization in the region of the central left internal jugular vein by at least four venous collateral vessels (arrows) is consistent with a venous occlusion.

 

Figure 3C
View larger version (106K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3c: Time-resolved coronal MIPs show marked venous collateral vessels in a patient with a left internal jugular vein occlusion. (a) Image acquired during the early arterial phase of contrast material administration shows predominantly pulmonary and systemic arterial opacification. Some residual contrast material from the injection can be seen in the right subclavian vein. (b) Image acquired in the early venous phase shows normal opacification of the right internal jugular, subclavian, and brachiocephalic veins but with abrupt tapering of the caliber of the left internal jugular vein (*). (c) Marked collateralization in the region of the central left internal jugular vein by at least four venous collateral vessels (arrows) is consistent with a venous occlusion.

 

Figure 4A
View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a: Comparison of (a–c) high-spatial-resolution images with (d, e) time-resolved coronal MIP images in a patient with superior vena cava occlusion. (a) A patent brachiocephalic vein (arrowhead) and a narrowed superior vena cava (*) are visible. (b, c) Additional coronal images show an enlarged azygous vein (thick arrow). (d) MIP image acquired at an early time point shows contrast material inflow via the right subclavian and cephalic veins to the right brachiocephalic vein (arrowhead) and into the peripheral superior vena cava (*), which appears to terminate abruptly. The azygous vein (arrows) is opacified and enlarged. The midportion of the azygous vein has been omitted from the imaging field of view and thus causes the appearance of a pseudo-obstruction. The right side of the heart does not demonstrate opacification; this finding is consistent with collateral flow around a completely occluded central superior vena cava. (e) Subsequent MIP image acquired 4 seconds after d shows contrast material in the inferior vena cava (arrow), with opacification of the right side of the heart and the main pulmonary artery.

 

Figure 4B
View larger version (133K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b: Comparison of (a–c) high-spatial-resolution images with (d, e) time-resolved coronal MIP images in a patient with superior vena cava occlusion. (a) A patent brachiocephalic vein (arrowhead) and a narrowed superior vena cava (*) are visible. (b, c) Additional coronal images show an enlarged azygous vein (thick arrow). (d) MIP image acquired at an early time point shows contrast material inflow via the right subclavian and cephalic veins to the right brachiocephalic vein (arrowhead) and into the peripheral superior vena cava (*), which appears to terminate abruptly. The azygous vein (arrows) is opacified and enlarged. The midportion of the azygous vein has been omitted from the imaging field of view and thus causes the appearance of a pseudo-obstruction. The right side of the heart does not demonstrate opacification; this finding is consistent with collateral flow around a completely occluded central superior vena cava. (e) Subsequent MIP image acquired 4 seconds after d shows contrast material in the inferior vena cava (arrow), with opacification of the right side of the heart and the main pulmonary artery.

 

Figure 4C
View larger version (135K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4c: Comparison of (a–c) high-spatial-resolution images with (d, e) time-resolved coronal MIP images in a patient with superior vena cava occlusion. (a) A patent brachiocephalic vein (arrowhead) and a narrowed superior vena cava (*) are visible. (b, c) Additional coronal images show an enlarged azygous vein (thick arrow). (d) MIP image acquired at an early time point shows contrast material inflow via the right subclavian and cephalic veins to the right brachiocephalic vein (arrowhead) and into the peripheral superior vena cava (*), which appears to terminate abruptly. The azygous vein (arrows) is opacified and enlarged. The midportion of the azygous vein has been omitted from the imaging field of view and thus causes the appearance of a pseudo-obstruction. The right side of the heart does not demonstrate opacification; this finding is consistent with collateral flow around a completely occluded central superior vena cava. (e) Subsequent MIP image acquired 4 seconds after d shows contrast material in the inferior vena cava (arrow), with opacification of the right side of the heart and the main pulmonary artery.

 

Figure 4D
View larger version (87K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4d: Comparison of (a–c) high-spatial-resolution images with (d, e) time-resolved coronal MIP images in a patient with superior vena cava occlusion. (a) A patent brachiocephalic vein (arrowhead) and a narrowed superior vena cava (*) are visible. (b, c) Additional coronal images show an enlarged azygous vein (thick arrow). (d) MIP image acquired at an early time point shows contrast material inflow via the right subclavian and cephalic veins to the right brachiocephalic vein (arrowhead) and into the peripheral superior vena cava (*), which appears to terminate abruptly. The azygous vein (arrows) is opacified and enlarged. The midportion of the azygous vein has been omitted from the imaging field of view and thus causes the appearance of a pseudo-obstruction. The right side of the heart does not demonstrate opacification; this finding is consistent with collateral flow around a completely occluded central superior vena cava. (e) Subsequent MIP image acquired 4 seconds after d shows contrast material in the inferior vena cava (arrow), with opacification of the right side of the heart and the main pulmonary artery.

 

Figure 4E
View larger version (98K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4e: Comparison of (a–c) high-spatial-resolution images with (d, e) time-resolved coronal MIP images in a patient with superior vena cava occlusion. (a) A patent brachiocephalic vein (arrowhead) and a narrowed superior vena cava (*) are visible. (b, c) Additional coronal images show an enlarged azygous vein (thick arrow). (d) MIP image acquired at an early time point shows contrast material inflow via the right subclavian and cephalic veins to the right brachiocephalic vein (arrowhead) and into the peripheral superior vena cava (*), which appears to terminate abruptly. The azygous vein (arrows) is opacified and enlarged. The midportion of the azygous vein has been omitted from the imaging field of view and thus causes the appearance of a pseudo-obstruction. The right side of the heart does not demonstrate opacification; this finding is consistent with collateral flow around a completely occluded central superior vena cava. (e) Subsequent MIP image acquired 4 seconds after d shows contrast material in the inferior vena cava (arrow), with opacification of the right side of the heart and the main pulmonary artery.

 
Time for Study Interpretation
The overall average interpretation time per session was 3.0 minutes ± 0.2 (standard deviation) for session 1 (time-resolved MIP images only), 3.4 minutes ± 0.2 for session 2 (high-spatial-resolution images only), and 3.3 minutes ± 0.2 for session 3 (both data sets) (Table 2). The interpretation time for time-resolved MIP images was significantly shorter than that for high-spatial-resolution data sets or both data sets; however, the difference between interpretation times for sessions 2 and 3 was not significant. The average interpretation time was less than 4 minutes for each session, regardless of radiologist experience.


View this table:
[in this window]
[in a new window]

  		Table 2. Mean Interpretation Times per Study

 
Determination of Side of Injection
Intravenous contrast material was administered via a right-sided peripheral intravenous catheter (n = 15), a left-sided intravenous catheter (n = 9), or a central venous catheter (n = 3). When time-resolved MIP images were available, readers accurately determined the correct side of contrast material injection in 23 (86%) and 22 (83%) patients in sessions 1 and 3, respectively; however, when time-resolved MIP images were not available (session 2), readers accurately determined the correct side of contrast material injection in only nine (32%) patients (P < .001 for comparison between session 2 and both session 1 and session 3).

Detection of Stenoses and Occlusions
Overall sensitivity for the detection of stenoses or occlusions was high, without a significant difference between session 1, 2, or 3 (0.95, 0.92, and 0.93, respectively; Table 3). However, interpretation of time-resolved MIP images alone resulted in significantly higher sensitivity in the detection of occlusions compared with that in sessions 2 and 3; this was particularly evident in the subset of resident readers (P < .05).


View this table:
[in this window]
[in a new window]

  		Table 3. Sensitivities and Specificities for Detecting Stenoses or Occlusions of the Central Veins of the Chest

 
Overall specificity for detecting stenoses or occlusions with time-resolved MIP images alone was 0.56. Specificity was significantly higher with high-spatial-resolution images alone (0.86, P < .001), and it was not further improved with the addition of time-resolved MIP images (0.84). Overall specificity for the detection of occlusions was 0.72 for session 1. Again, the availability of high-spatial-resolution images significantly increased the specificity to 0.97 for both session 2 and session 3 (P < .001). However, for the faculty reader subgroup, the addition of time-resolved MIP images resulted in a small but significant improvement in specificity for the detection of occlusions between sessions 2 and 3 (0.96 and 0.99, respectively; P = .03).

The overall interreader agreement was moderate, with {kappa} values of 0.50–0.55 (Table 4). For the faculty reader subset, interreader agreement was substantial, with concordance of 96.3% and {kappa} values of 0.60–0.70.


View this table:
[in this window]
[in a new window]

  		Table 4. Interobserver Agreement for Rating Central Venous Stenoses or Occlusions

 
Reader Confidence Level
Interpretation of time-resolved MIP images alone resulted in the lowest overall average confidence level, with an average confidence level of 2.6 on a scale of 1–4 (4 representing the highest level of confidence). The average confidence level for interpretation of the high-spatial-resolution data sets was significantly higher at 3.4 (P < .001). The addition of time-resolved MIP images to the high-spatial-resolution images improved the confidence level slightly further, to 3.5 (session 2 vs session 3, P < .001).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
One of the primary advantages of time-resolved MIP imaging is that it enables dynamic visualization of venous blood flow in a manner similar to that in conventional angiography, which facilitates visualization of the enlarged collateral veins that often accompany chronic venous stenoses and occlusions. In general, time-resolved techniques have been shown to be sensitive and specific as stand-alone sequences when they are used to evaluate midcaliber high-flow vessels, such as the renal arteries (20) and arteriovenous fistulas and grafts (14,21). Sensitivity and specificity in the detection of stenoses and occlusions in smaller high-flow vessels, such as the peripheral arteries, are not quite as high (22,23). To our knowledge, use of time-resolved imaging in the evaluation of large slow-flowing veins has not been reported; therefore, we sought to evaluate the central veins of the chest.

The overall average interpretation time for time-resolved MIP images alone was significantly shorter than that for high-spatial-resolution images or for both data sets. This is not surprising when one considers that the time-resolved data sets were composed of only 15 MIP images that could be viewed quickly in succession, as if one were viewing a conventional venogram. On the other hand, the multiple contiguous high-spatial-resolution coronal images required one to scroll through approximately 112 images per time point to locate the veins of interest, and there were three time points per patient. It was interesting that there was no significant increase in interpretation time when we read both data sets together when compared with the interpretation time when we read the high-spatial-resolution data set alone, despite the fact that there were more images to interpret. This was likely secondary to interpretation strategies, such as reviewing one data set comprehensively, with a limited review of the second data set to confirm the findings. Another possible explanation is that a decreased amount of time was spent deliberating on any ambiguous venous segments by virtue of having two different image sets. For the junior radiologists, the experience provided by the previous two reading sessions also may have had an effect.

There were three primary advantages to adding time-resolved MIP images to the traditionally used high-spatial-resolution images. First, there was a small but significant increase in reader confidence. Second, there was a small but significant improvement in specificity in the detection of occlusions among the senior readers. Since time-resolved MIP images are superior to static high-spatial-resolution data sets for dynamic depiction of collateral vessels, the benefit likely resulted from the inference of severe stenoses and occlusions when marked collateral vessels were depicted. Third, the use of time-resolved MIP images resulted in a marked improvement in our ability to determine the side of contrast material injection. This is of clinical importance, as nondiluted contrast material occasionally becomes trapped in the venous valves within the subclavian veins, causing potential image degradation in adjacent vessels because of T2*-related susceptibility artifacts (24,25).

Interpretation of time-resolved MIP images alone resulted in high sensitivity for the detection of stenoses or occlusions, as well as for the detection of occlusions only (0.95 and 0.88, respectively); however, specificity was poor (0.56 for stenoses or occlusions and 0.72 for occlusions only). These findings imply that if the central veins are patent without stenosis, the sensitivity and specificity will be excellent; however, when stenosis or occlusion is present, it might be difficult to determine which veins are abnormal and to what extent, likely because of extremely slow contrast material flow into severely diseased vein segments. Thus, interpretation of time-resolved MIP images is inadequate as a stand-alone technique in the evaluation of the central venous system for stenosis or occlusion. However, because of its high sensitivity and low specificity, interpretation of time-resolved MIP images alone qualifies as an effective screening technique for use in patients with a low pretest probability of having severe venous stenoses or occlusions. Additional characteristics that support the use of this approach as a screening technique include a lower contrast material dose requirement (10 mL vs 30 mL), a shorter acquisition time (2 minutes vs approximately 5 minutes), and a significantly shorter time for image interpretation compared with the time needed to interpret high-spatial-resolution data sets. Another approach could entail a sequence protocol in which time-resolved data sets are acquired initially, with a subsequent immediate review for any questionable findings to determine whether an additional contrast material load and a more time-consuming acquisition of high-spatial-resolution images are necessary.

The overall quantity of gadolinium administered is of particular importance since gadolinium-based contrast agents recently have been implicated as a causative factor in the development of nephrogenic systemic fibrosis in patients with renal insufficiency or failure in a dose-dependent manner (7–11). On the basis of these results, the Food and Drug Administration issued a public health advisory concerning the use of high-dose gadolinium-containing contrast agents in patients with advanced renal failure (26). Patients with renal failure are particularly prone to developing central venous disease (27). In fact, in this study, 12 (44%) of the 27 patients referred for evaluation of the central veins had renal failure or insufficiency and thus constituted a population at risk for subsequent development of nephrogenic systemic fibrosis. The normal gadolinium-based contrast agent dose used at our institution to obtain high-spatial-resolution images for assessment of central veins of the chest is 30 mL, whereas only 10 mL of contrast material is used to obtain time-resolved data sets. Thus, study termination for normal or nearly normal time-resolved data sets, as suggested earlier in this article could serve to limit the gadolinium dose by 75%. If this strategy had been implemented at the time of initial data acquisition, four (15%) of the 27 patients may have received only a fraction (25%) of the regular contrast agent dose.

The radiologists who participated in this study had widely varying levels of experience. Overall, the junior readers achieved sensitivity and specificity for the detection of stenoses and occlusions similar to those achieved by the faculty readers. The average interpretation times were also similar. These findings suggest that MR venograms of the chest can be competently interpreted by radiologists with limited experience.

One limitation of this study was that we used the consensus reading performed by the two senior readers when conventional venograms were not available for reference values instead of obtaining conventional venograms for all patients. However, the faculty readings were concordant for the majority (96%) of cases; therefore, consensus reading rarely was required. Furthermore, use of consensus reading was deemed acceptable since the aim of this study was not to determine the general sensitivity and specificity of MR data sets, as compared with conventional venography data sets, but rather to elucidate the relative degree of diagnostic information gleaned and the efficiency in interpretation achieved with time-resolved MIP images. Additionally, conventional angiography has substantial limitations, such as a routine failure to depict the internal jugular veins and the required use of nephrotoxic contrast agents that are associated with a substantial incidence of allergic reactions (1).

Another limitation of our study was the use of MR systems with different magnetic field strengths. However, the MR acquisition protocols for each unit were individually optimized for visualization of the central veins of the chest. Furthermore, vessel image quality has been shown to be highly comparable at 1.5 T and 3.0 T (24,28–30).

A third limitation of our study was our use of coronal MIP images without the coronal source images for the time-resolved data set and our use of coronal source images without the reconstructed MIP images for the high-spatial-resolution data set. However, the choice of images available for interpretation was based on current standard practices for MR venogram interpretation.

In summary, the addition of time-resolved angiographic images to the traditionally used static high-spatial-resolution contrast-enhanced data sets resulted in a significant improvement in specificity for the detection of occlusions of the central veins of the chest for the faculty readers, probably because of improved depiction of collateral vessels, without increasing the time needed for study interpretation. Furthermore, there was a significant improvement in confidence and ability to determine the side of contrast material injection for all readers. Use of time-resolved MR angiography as a stand-alone technique demonstrated high sensitivity but inadequate specificity in the assessment of the central veins. However, we believe that the high sensitivity, rapid acquisition time, and need for only a fraction of the overall gadolinium dose should lead to the use of time-resolved MR angiography as an initial evaluation sequence, with the potential to obviate more time-consuming high-spatial-resolution imaging and additional gadolinium-based contrast agent administration.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
IMPLICATIONS FOR PATIENT CARE
 References
 


   IMPLICATIONS FOR PATIENT CARE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
References
 


   ACKNOWLEDGMENTS
 
We thank Siemens Medical Solutions for providing the non–commercially available time-resolved echo-shared angiographic technique sequence free of charge.


   FOOTNOTES
 

Abbreviations: MIP = maximum intensity projection

See Materials and Methods for pertinent disclosures.

Author contributions: Guarantors of integrity of entire study, C.Y.K., E.M.M.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, C.Y.K., E.M.M.; clinical studies, C.Y.K., R.A.M., J.A.B., E.D.W., C.E.S., E.M.M.; statistical analysis, C.Y.K., D.M.D.; and manuscript editing, C.Y.K., J.A.B., D.M.D., C.E.S., E.M.M.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 

  1. Johnsrude IS. Venography and lymphangiography. In: Johnsrude IS, Jackson DC, Dunnick NR, eds. A practical approach to angiography. 2nd ed. Boston, Mass: Little, Brown, 1987; 565–650.
  2. Nael K, Laub G, Finn JP. Three-dimensional contrast-enhanced MR angiography of the thoraco-abdominal vessels. Magn Reson Imaging Clin N Am 2005;13(2):359–380.[CrossRef][Medline]
  3. Thornton MJ, Ryan R, Varghese JC, Farrell MA, Lucey B, Lee MJ. A three-dimensional gadolinium-enhanced MR venography technique for imaging central veins. AJR Am J Roentgenol 1999;173:999–1003.[Abstract/Free Full Text]
  4. Tanju S, Sancak T, Dusunceli E, Yagmurlu B, Erden I, Sanlidilek U. Direct contrast-enhanced 3D MR venography evaluation of upper extremity deep venous system. Diagn Interv Radiol 2006;12:74–79.[Medline]
  5. Shinde TS, Lee VS, Rofsky NM, Krinsky GA, Weinreb JC. Three-dimensional gadolinium-enhanced MR venographic evaluation of patency of central veins in the thorax: initial experience. Radiology 1999;213:555–560.[Abstract/Free Full Text]
  6. Oxtoby JW, Widjaja E, Gibson KM, Uzoka K. 3D gadolinium-enhanced MRI venography: evaluation of central chest veins and impact on patient management. Clin Radiol 2001;56:887–894.[CrossRef][Medline]
  7. Sadowski EA, Bennett LK, Chan MR, et al. Nephrogenic systemic fibrosis: risk factors and incidence estimation. Radiology 2007;243(1):148–157.[Abstract/Free Full Text]
  8. Broome DR, Girguis MS, Baron PW, Cottrell AC, Kjellin I, Kirk GA. Gadodiamide-associated nephrogenic systemic fibrosis: why radiologists should be concerned. AJR Am J Roentgenol 2007;188:586–592.[Abstract/Free Full Text]
  9. Kuo PH, Kanal E, Abu-Alfa AK, Cowper SE. Gadolinium-based MR contrast agents and nephrogenic systemic fibrosis. Radiology 2007;242(3):647–649.[Free Full Text]
  10. High WA, Ayers RA, Chandler J, Zito G, Cowper SE. Gadolinium is detectable within the tissue of patients with nephrogenic systemic fibrosis. J Am Acad Dermatol 2007;56(1):21–26.[CrossRef][Medline]
  11. Marckmann P, Skoy L, Rossen K, et al. Nephrogenic systemic fibrosis: suspected causative role of gadodiamide used for contrast-enhanced magnetic resonance imaging. J Am Soc Nephrol 2006;17(9):2359–2362.[Abstract/Free Full Text]
  12. Korosec FR, Frayne R, Grist TM, Mistretta CA. Time-resolved contrast-enhanced 3D MR angiography. Magn Reson Med 1996;36(3):345–351.[Medline]
  13. Du J, Thornton FJ, Mistretta CA, Grist TM. Dynamic MR venography: an intrinsic benefit of time-resolved MR angiography. J Magn Reson Imaging 2006;24:922–927.[CrossRef][Medline]
  14. Pinto C, Hickey R, Carroll TJ, et al. Time-resolved MR angiography with generalized autocalibrating partially parallel acquisition and time-resolved echo-sharing angiographic technique for hemodialysis arteriovenous fistulas and grafts. J Vasc Interv Radiol 2006;17(6):1003–1009.[CrossRef][Medline]
  15. Michaely HJ, Herrmann KA, Dietrich O, Reiser MF, Schoenberg SO. Quantitative and qualitative characterization of vascularization and hemodynamics in head and neck tumors with a 3D magnetic resonance time-resolved echo-shared angiographic technique (TREAT): initial results. Eur Radiol 2007;17(4):1101–1110.[CrossRef][Medline]
  16. Michaely HJ, Nael K, Schoenberg SO, et al. Renal perfusion: comparison of saturation-recovery TurboFLASH measurements at 1.5T with saturation-recovery TurboFLASH and time-resolved echo-shared angiographic technique (TREAT) at 3.0T. J Magn Reson Imaging 2006;24(6):1413–1419.[CrossRef][Medline]
  17. Brauck K, Maderwald S, Vogt FM, Zenge M, Barkhausen J, Herborn CU. Time-resolved contrast-enhanced magnetic resonance angiography of the hand with parallel imaging and view sharing: initial experience. Eur Radiol 2007;17(1):183–192.[CrossRef][Medline]
  18. Fink C, Puderbach M, Ley S, Zaporozhan J, Plathow C, Kauczor HU. Time-resolved echo-shared parallel MRA of the lung: observer preference study of image quality in comparison with non-echo-shared sequences. Eur Radiol 2005;15(10):2070–2074.[CrossRef][Medline]
  19. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33(1):159–174.[CrossRef][Medline]
  20. Masunaga H, Takehara Y, Isoda H, et al. Assessment of gadolinium-enhanced time-resolved three-dimensional MR angiography for evaluating renal artery stenosis. AJR Am J Roentgenol 2001;176:1213–1219.[Abstract/Free Full Text]
  21. Froger CL, Duijm LE, Liem YS, et al. Stenosis detection with MR angiography and digital subtraction angiography in dysfunctional hemodialysis access fistulas and grafts. Radiology 2005;234:284–291.[Abstract/Free Full Text]
  22. Zhang HL, Khilnani NM, Prince MR, et al. Diagnostic accuracy of time-resolved 2D projection MR angiography for symptomatic infrapopliteal arterial occlusive disease. AJR Am J Roentgenol 2005;184:938–947.[Abstract/Free Full Text]
  23. Swan JS, Carroll TJ, Kennell TW, et al. Time-resolved three dimensional contrast-enhanced MR angiography of the peripheral vessels. Radiology 2002;225(1):43–52.[Abstract/Free Full Text]
  24. Lee VS, Martin DJ, Krinsky GA, Rofsky NM. Gadolinium-enhanced MR angiography: artifacts and pitfalls. AJR Am J Roentgenol 2000;175:197–205.[Free Full Text]
  25. Lee YJ, Chung TS, Joo JY, Chien D, Laub G. Suboptimal contrast-enhanced carotid MR angiography from the left brachiocephalic venous stasis. J Magn Reson Imaging 1999;10:503–509.[CrossRef][Medline]
  26. Public health advisory: update on magnetic resonance imaging (MRI) contrast agents containing gadolinium and nephrogenic fibrosing dermopathy. U.S. Food and Drug Administration Web site. http://www.fda.gov/cder/drug/advisory/gadolinium_agents_20061222.htm. Published December 22, 2006. Accessed May 3, 2007.
  27. MacRae JM, Ahmed A, Johnson N, Levin A, Kiaii M. Central vein stenosis: a common problem in patients on hemodialysis. ASAIO J 2005;51(1):77–81.[CrossRef][Medline]
  28. Willinek WA, Bayer T, Gieseke J, et al. High spatial resolution contrast-enhanced MR angiography of the supraaortic arteries using the quadrature body coil at 3.0T: a feasibility study. Eur Radiol 2007;17(3):618–625.[CrossRef][Medline]
  29. Frydrychowicz A, Bley TA, Winterer JT, et al. Accelerated time-resolved 3D contrast-enhanced MR angiography at 3T: clinical experience in 31 patients. MAGMA 2006;19(4):187–195.[CrossRef][Medline]
  30. Cashen TA, Carr JC, Shin W, et al. Intracranial time-resolved contrast-enhanced MR angiography at 3T. AJNR Am J Neuroradiol 2006;27(4):822–829.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
C. Y. Kim and E. M. Merkle
Time-Resolved MR Angiography of the Central Veins of the Chest
Am. J. Roentgenol., November 1, 2008; 191(5): 1581 - 1588.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2472070790v1
247/2/558    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kim, C. Y.
Right arrow Articles by Merkle, E. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kim, C. Y.
Right arrow Articles by Merkle, E. M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE