| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, University of California-San Francisco, 521 Parnassus
Ave., Rm. C-324C, Box 0628, San Francisco, CA 94143-0628.
2 Present address: Department of Radiology, University of Chicago Hospitals,
Chicago, IL.
3 Department of Medicine, University of California-San Francisco, San Francisco,
CA.
Received February 3, 2004;
accepted after revision April 3, 2004.
Address correspondence to B. M. Yeh
(benyeh{at}itsa.ucsf.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. We retrospectively identified 127 patients
who underwent contemporaneous contrast-enhanced CT of the chest or abdomen and
echocardiography. On CT, the presence of retrograde inferior vena cava or
hepatic vein opacification and the rate of IV contrast injection (> 3
mL/sec, high; 
3 mL/sec, low) were recorded. On echocardiography, the
presence of tricuspid regurgitation, pulmonary hypertension, or right
ventricular systolic dysfunction was recorded.
RESULTS. Retrograde inferior vena cava or hepatic vein opacification was more common in studies performed with a high rather than a low contrast injection rate (28/56 vs 6/71 patients, respectively; p < 0.01). This finding was 31% sensitive (5/16) and 98% specific (54/55) for right-sided heart disease at low contrast injection rates, and 81% sensitive (17/21) and 69% specific (24/35) at high injection rates. Multivariate logistic regression models showed that high injection rate, tricuspid regurgitation, pulmonary hypertension, and right ventricular systolic dysfunction were independent predictors of retrograde inferior vena cava or hepatic vein opacification (p < 0.05 for each).
CONCLUSION. Retrograde opacification of the inferior vena cava or hepatic veins on CT is a specific but insensitive sign of right-sided heart disease at low contrast injection rates, but the usefulness of this classic sign decreases with high injection rates. This realization is important because many centers increasingly use high-injection-rate CT.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Echocardiography Technique
Transthoracic echocardiography was performed in 109 patients (44 men and 65
women), and transesophageal echocardiography was performed in 18 patients (10
men and eight women) according to a standard protocol for accepted clinical
indications. Transthoracic echocardiography was performed with 3.5- and
4.0-MHz transducers (Siemens Acuson Sequoia or Phillips Sonos 5500).
Transesophageal echocardiography was performed with 7.0-MHz transducers
(Acuson Sequoia). One author retrieved the echocardiography reports and
tabulated the following results: ejection fraction; the presence of tricuspid
valve regurgitation, pulmonary hypertension, right ventricular systolic
dysfunction, right ventricular hypertrophy, pulmonic valve regurgitation, left
ventricular enlargement, right atrial diastolic collapse (pretamponade),
mitral valve regurgitation, mitral valve stenosis, aortic valve regurgitation,
and aortic valve stenosis; and arrhythmia. Valvular regurgitation was recorded
when regurgitation was reported as mild to moderate or greater. Pulmonary
hypertension was considered to be present when pulmonary artery systolic
pressures were estimated to be greater than 35 mm Hg
[6].
CT Technique
The CT technique is described here separately for patients who received IV
contrast material at rates of injection greater than 3 mL/sec (high rate of
injection) and for patients who received IV contrast material at rates of
injection equal to or less than 3 mL/sec (low rate of injection). All patients
received IV contrast material through peripheral venous catheters in the arms.
Fifty-six examinations were performed on helical CT scanners (LightSpeed Plus,
n = 36; LightSpeed QX/I, n = 20; GE Healthcare) after the IV
administration of 100 mL (n = 37) or 150 mL (n = 19) of
iohexol (Omnipaque 350, Nycomed Amersham) at a high rate of injection (> 3
mL/sec). Thirty-seven of these examinations were performed in women and 19
were performed in men. Indications and additional technical parameters for
these scans were suspected pulmonary embolism (n = 37), slice
collimation of 1.25 mm and scanning delay of 20 sec; suspected aortic injury
(n = 6), slice collimation of 2.5 mm and scanning delay of 20 sec;
and multiphase evaluation of the abdomen for possible malignancy (n =
10) or before organ donation (n = 3), slice collimation of 5 mm and
earliest scanning delays of 20 sec (n = 3), 40 sec (n = 7),
or 60 sec (n = 3).
The other 71 examinations were performed using helical CT scanners (LightSpeed QX/I, n = 43; LightSpeed Plus, n = 17; HiSpeed CT/I, n = 1; HiSpeed Advantage, n = 10; GE Healthcare) with slice collimation of 5 mm and IV injection of 100 mL (n = 7) or 150 mL (n = 64) of iohexol (Omnipaque 350) at a rate of 1-3 mL/sec and earliest scanning delay of 45 sec (n = 7), 70 sec (n = 61), or 180 sec (n = 2). The indications for these CT scans were evaluation of possible malignancy (n = 30), sepsis or fluid collections (n = 23), and abdominal pain (n = 18). Thirty-eight of these examinations were performed in women and 33 were performed in men.
Image Interpretation
One attending radiologist with subspecialty training in abdominal imaging
who was unaware of echocardiography results and patient history reviewed the
earliest phase of contrast enhancement of each CT examination. Retrograde
opacification of the hepatic veins or inferior vena cava was considered to be
present when the attenuation of the lumen of the hepatic veins or inferior
vena cava was visibly higher than that of a more caudal segment of hepatic
vein or inferior vena cava. The smallest maximal axial diameter of the
inferior vena cava was measured above the level of the liver (suprahepatic)
and at the level of the confluence of the hepatic veins (intrahepatic). The
short-axis diameter of the largest right, middle, and left hepatic veins was
measured within 2 cm of the inferior vena cava. The intrahepatic inferior vena
cava was not seen well enough to measure in six (5%) of 127 patients, and the
hepatic veins were not seen well enough to measure in 17 (14%) of 127
patients.
The presence or absence of ascites, defined as intraperitoneal fluid attenuation between the liver or spleen and the diaphragm, was also recorded. Ascites elsewhere in the abdomen and pelvis were ignored because CT examinations performed only through the chest did not extend below the level of the adrenal glands. For patients in whom the chest was imaged on CT, whether the right or the left arm was injected with IV contrast material was recorded as follows: If much more intense contrast was seen in the left or right axillary or subclavian veins than in the contralateral axillary or subclavian vein, then the ipsilateral side was recorded as the arm that was injected with IV contrast material; if the axillary or subclavian veins were not imaged or no obvious high-intensity contrast was seen in an upper extremity vein, then the side of injection was recorded as indeterminate.
Statistical Analysis
Statistical analysis was performed using Stata version 7.0 software package
(Stata). The frequencies of retrograde opacification of the inferior vena cava
or hepatic vein in examinations performed with high and low rates of contrast
injection were compared using chi-square analysis. The sensitivity,
specificity, positive predictive value, and negative predictive value of
retrograde opacification into the inferior vena cava or hepatic veins for
detecting right-sided heart disease were calculated. Right-sided heart disease
was defined as the echocardiographic finding of tricuspid regurgitation,
pulmonary hypertension, or right ventricular systolic dysfunction. Univariate
analysis using the chi-square or Fisher's exact test as appropriate
[7] was performed to assess the
strength of individual echocardiographic and clinical determinants of
retrograde IV contrast reflux into the inferior vena cava or hepatic veins. To
assess the independence of determinants of retrograde IV contrast reflux into
the inferior vena cava or hepatic vein, multivariate logistic regression was
performed. All multivariate analyses accounted for variations in both the
volume of IV contrast material that was administered and the scanning delay
after contrast material injection. Multivariate logistic regression was
performed using a backward selection method to decrease the likelihood of
excluding negatively confounding factors
[8,
9].
Differences in the mean diameter of venous structures between patients with and without right-sided heart disease were compared using box plots and Student's t tests. Mean hepatic vein diameter was the mean of the right, middle, and left hepatic veins short-axis diameters. For all tests, a p value less than 0.05 was considered to be significant.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
The finding of diastolic collapse (pretamponade) on echocardiography was found in only two patients, neither of whom had retrograde reflux on CT. The mean diameters of the inferior vena cava and hepatic veins on CT were larger in patients with right-sided cardiac disease than in those without; however, substantial overlap of venous diameters was seen between the two groups (Fig. 2).
|
Univariate analysis showed that high injection rate, tricuspid regurgitation, pulmonary hypertension, right ventricular systolic dysfunction, and female sex were predictive of the finding of contrast reflux into the inferior vena cava or hepatic veins on CT (Table 2). However, on multivariate analysis, only high injection rate, tricuspid regurgitation, pulmonary hypertension, and right ventricular systolic dysfunction were found to be independent predictors of retrograde inferior vena cava or hepatic vein opacification on CT (p < 0.001, < 0.01, = 0.05, and < 0.005, respectively). A trend toward statistical significance was seen for female sex as a predictor for retrograde contrast reflux in the final multivariate logistic regression model (p = 0.055). No significant difference was seen in the prevalence of tricuspid regurgitation, pulmonary hypertension, or right ventricular systolic dysfunction among the women (10/75, 18/75, and 5/75 women, respectively) compared with the men (3/52, 6/52, and 7/52 men, respectively, p > 0.05 for all comparisons) in our population. Arrhythmia and left-sided cardiac disease, such as ejection fraction and mitral valve regurgitation, were not found to be predictive of IV contrast material reflux into the inferior vena cava or hepatic veins.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our study showed that, with IV contrast injection rates less than 3 mL/sec, the finding of retrograde IV contrast reflux into the inferior vena cava or hepatic veins on CT is highly specific (54/55 patients [98%]) but not sensitive (5/16 patients [31%]) for right-sided heart disease. The sensitivity of this sign increases to 81% (17/21 patients) at IV contrast injection rates greater than 3 mL/sec, but the specificity drops to 69% (24/35 patients). To our knowledge, the contribution of IV injection rate to the finding of retrograde opacification of the inferior vena cava or hepatic veins has not been previously described. The decreased specificity of finding retrograde opacification of the inferior vena cava or hepatic veins as a sign of right-sided heart disease on CT scans obtained with high-contrast injection rate is important to recognize because many centers are now performing an increasing proportion of CT examinations with high injection rates.
The higher sensitivity and decreased specificity for detecting heart disease at high IV contrast injection rates compared with low injection rates may be due to several possible reasons. One explanation is that retrograde flow of contrast material initially occurs regardless of injection rate, but because the high-injection-rate examinations were performed with shorter scanning delays on average than the low-injection-rate examinations, the high-injection-rate group (i.e., the shorter scanning delay group) may have been imaged before blood flowing antegrade from the inferior vena cava and hepatic veins could push the refluxed contrast material back into the right ventricle. Furthermore, because administration of IV contrast material increases intravascular volume [11], patients who are given a larger bolus of IV contrast material may also have a larger volume of recirculated blood flowing antegrade to push refluxed contrast material back into the heart. However, our multivariate analysis showed that even when accounting for scanning delay and volume of injected IV contrast material, the injection rate remained an independent determinant of retrograde opacification of the inferior vena cava or hepatic veins. This analysis suggests that the rate of contrast injection itself contributed to the frequency of finding reflux and that a faster injection rate acts as a volume challenge that unmasks decreased right cardiac output reserve.
In our multivariate analysis, we found that several factors independently increase the likelihood of retrograde opacification of the inferior vena cava and hepatic veins on CT: namely, right ventricular systolic dysfunction, high IV contrast injection rate, tricuspid regurgitation, and pulmonary hypertension, the more likely contrast reflux will be seen.
Another factor that may contribute to finding retrograde flow of contrast material in the inferior vena cava or hepatic veins is female sex. Female sex was significantly associated with retrograde reflux on univariate analysis (p < 0.05) and was nearly significantly associated with retrograde reflux on multivariate analysis (p = 0.055). This finding may be simply due to the smaller average body mass of female patients compared with male patients. A smaller body mass and intravascular volume is likely to experience greater intravascular shifts for a given injection rate than would a larger body mass and intravascular volume. Alternative possibilities include differences between men and women in autonomic response of the cardiovascular system [12-14], in compliance of small vessels [15], or in disease prevalence beyond that measured in our study.
Although we attempted to include as many patient variables as possible in our study, our results have several limitations. First, patients in our study population probably have a higher prevalence of cardiac disease than the general population because all patients in this study underwent clinically indicated IV contrast-enhanced CT of the chest or abdomen within 14 days of having clinical echocardiography. However, most of our patients (90/127) were not shown to have right-sided heart disease on echocardiography; these patients acted as a control group in the univariate analysis and allowed us to perform a more robust multivariate analysis than could be previously performed. Although retrospective comparison with a different group of case-matched patients without echocardiography could have been performed, it would be difficult to prove without echocardiography that such patients did not indeed have occult heart disease. Second, all patients were imaged at a single tertiary care hospital in North America, which may limit the extrapolation of our results to other institutions. Third, we could not readily determine the intravascular volume of the patients or the time interval from a patient's last meal to CT. These factors are known to affect direction of portal venous flow in the liver [16-18] and likely affect antegrade hepatic venous flow and the likelihood of retrograde reflux from the right atrium on IV contrast-enhanced CT.
Undoubtedly, these factors varied among patients but would have acted equally among patients both with and without right-sided heart disease and for CT examinations performed with both high and low contrast injection rates and are likely not significant confounding factors. Fourth, echocardiography results in our study were obtained from the echocardiography reports rather than reinterpretation of the source images, and therefore, observer variability may be present. However, such variability may reflect the variability in reporting in the more general clinical setting, and the variability would also have affected all patients regardless of whether they had cardiac disease or different CT contrast-injection rates.
Notwithstanding these limitations, we found that retrograde IV contrast material opacification of the inferior vena cava or hepatic veins on CT is a specific but insensitive sign of right-sided heart disease at low injection rates. However, the usefulness of this classic sign for suggesting right-sided heart disease is diminished with high injection rates, and this realization is important because many centers are increasing the use of high-injection-rate contrast-enhanced CT.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  R. Ryan, S. Abbara, R. R. Colen, S. Arnous, M. Quinn, R. C. Cury, and J. D. Dodd Cardiac Valve Disease: Spectrum of Findings on Cardiac 64-MDCT Am. J. Roentgenol., May 1, 2008; 190(5): W294 - W303. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Kandpal, R. Sharma, S. Gamangatti, D. N. Srivastava, and S. Vashisht Imaging the Inferior Vena Cava: A Road Less Traveled RadioGraphics, May 1, 2008; 28(3): 669 - 689. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Sheth and E. K. Fishman Imaging of the Inferior Vena Cava with MDCT Am. J. Roentgenol., November 1, 2007; 189(5): 1243 - 1251. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Castaner, X. Gallardo, J. Rimola, Y. Pallardo, J. M. Mata, J. Perendreu, C. Martin, and D. Gil Congenital and acquired pulmonary artery anomalies in the adult: radiologic overview. RadioGraphics, March 1, 2006; 26(2): 349 - 371. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Cobelli, M. Zompatori, E. Lombardo, N. Sverzellati, B. M. Yeh, and P. Kurzman Opacification of the Inferior Vena Cava and Inspiration-Associated Artifacts Am. J. Roentgenol., April 1, 2005; 184(4): 1361 - 1362. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
