| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 All authors: Department of Radiology, Duke University Medical Center, Box 3808, Rm. 2526, Blue Zone South, Durham, NC 27710.
Received October 26, 1999;
accepted after revision December 20, 1999.
Address correspondence to M. A. Kliewer.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Thirty-eight transjugular intrahepatic portosystemic shunts in 34 consecutive patients were studied using Doppler sonography. Peak velocities were measured in the mid shunt with the patient in three positions (supine, sitting upright, and left lateral decubitus) and two respiratory states (deep inspiration and quiet respiration). A mixed linear regression model was used to assess statistically significant differences among the six velocity measurements.
RESULTS. Peak velocities in the mid stent averaged 22 cm/sec greater in quiet respiration than in deep inspiration, which was a significant difference (p < 0.00001). Differences in velocities in the three patient positions were not significant (p = 0.53). Using 90-190 cm/sec as the normal range, the peak velocity shifted from normal to abnormal levels by changing respiratory state in 17 (45%) of 38 studies. Using 60 cm/sec as the lower normal limit, the peak velocity fell below the normal range with inspiration in 10 (26%) of 38 studies. In 12 (32%) of 38 studies, a decline in peak velocity exceeding 50 cm/sec could be induced by inspiration.
CONCLUSION. Peak systolic velocity in transjugular intrahepatic portosystemic shunts is substantially altered by the respiratory state of the patient at the time of the measurement, but not by the patient position. Respiratory state must be taken into account in the interpretation of peak velocity for shunt stenosis.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Obtaining angle-corrected velocity measurements in some TIPS can be problematic. These stents are often located deep in the abdomen and are oriented perpendicular to the transducer; consequently, the sonographer may have the patient change position or breathing level to optimize the Doppler angle and reveal flow in the stent lumen. However, changing the position and respiratory state of a patient could materially affect venous flow in the abdomen. Inspiration, expiration, and Valsalva's maneuver are known to alter venous flow in the inferior vena cava and extremities; the height of the right atrium relative to the systemic veins governs the hydrostatic pressure in the venous system [13]. Because the TIPS establishes a direct communication between the portal vein and the right atrium, we hypothesized that patient position and respiratory state could potentially influence venous flow through the stent. To our knowledge, there has been no accounting for these factors in the debate over normal and abnormal TIPS velocities. Therefore, the purpose of this study was to prospectively examine the effects of patient position and respiratory state on the measurement of Doppler velocities in TIPS stents.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Hepatic Doppler studies were performed using standard sonography equipment (HDI 3000 or 5000, Advanced Technology Laboratories, Bothell, WA; or Sequoia, Acuson, Mountain View, CA). Phased array transducers with center frequencies that ranged from 2.25 to 3.5 MHz were used. In addition to our standard TIPS protocol, which includes Doppler evaluation of the hepatic vasculature and TIPS, we obtained spectral Doppler waveforms in the middle portion of the stent with the patient supine, sitting upright, and in the left lateral decubitus position. In each of these positions, velocities were recorded with the patient in deep inspiration and in quiet respiration. The measured angle of insonation was less than 60° in all patients. The maximum flow velocity was recorded for each of the six combinations of two respiratory states and three positions. Technical parameters for Doppler imaging such as pulse repetition frequency, wall filter, Doppler gain, and transmitted frequency were optimized to enhance sensitivity to flow. All examinations were interpreted by an experienced sonologist.
Patients who had clinical or Doppler evidence of TIPS dysfunction underwent portal venography. Doppler evidence included peak velocities outside a 90-190 cm/sec range, a decline in intrashunt velocities greater than 50 cm/sec on sequential studies, portal vein velocities less than 30 cm/sec, hepatopetal flow within the left portal vein, and visible stenosis in the TIPS using color or power Doppler imaging. Portal venography was performed after access to the right internal jugular vein. An angiographic catheter was advanced into the inferior vena cava, hepatic vein, and through the shunt. Digital subtraction angiography was performed with the catheter positioned in the main portal vein. Portosystemic pressure gradients were recorded. Visible stenosis or abnormal portosystemic pressure (mean gradient > 12 mm Hg) was considered evidence of shunt dysfunction.
The maximum flow velocity in the mid stent in the six combinations of patient position and respiratory state were compared with one another using a mixed linear regression model appropriate for this randomized block study design. With this method, the statistical significance of patient position and respiratory state could be examined separately, and any interaction between the two variables could be shown. The clinical importance of those effects found to be statistically significant was assessed by comparing the velocity change with reported threshold values defining the normal velocity range in TIPS. The subset of patients with evidence of shunt dysfunction on portal venography was compared with those without evidence of shunt dysfunction to assess systematic differences in the velocity responses attributable to respiratory state and patient position. All statistical tests were considered significant at p values of 0.05 or less.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
The decline of velocity between respiratory states was statistically significant (p < 0.00001). The differences between the measured velocities in the three different patient positions were not significant (p = 0.53). Likewise, there was no significant interaction between respiratory state and patient position; no combination of the two variables was statistically significant (p = 0.54).
Velocity changes induced by respiratory state were compared with reported threshold values for the normal velocity range in TIPS. Using 90-190 cm/sec as the normal range [3,5], the peak velocity shifted from a normal level to an abnormal level in response to a change in respiratory state in 17 (45%) of 38 studies (Fig. 1A,1B). In the other 21 studies, changes in intrashunt velocities remained in the 90-190 cm/sec range, but decreased an average of 11 cm/sec in response to deep inspiration. If a more stringent threshold of 60 cm/sec was used [2,6], the peak velocity with deep inspiration decreased below normal in 10 (26%) of 38 studies (Fig. 2A,2B). In 12 (32%) of 38 studies, a decline in peak velocities exceeding 50 cm/sec could be induced in at least one position by deep inspiration alone.
|
|
|
|
Twelve patients underwent portal venography to evaluate shunt dysfunction suggested by clinical or Doppler evidence. Doppler studies showed intrashunt velocities of less than 90 cm/sec in seven cases, intrashunt velocities greater than 190 cm/sec in two patients, hepatopetal flow in the left portal vein in one patient, and a decline in intrashunt velocities greater than 50 cm/sec on sequential studies in two patients. Shunt dysfunction was identified in seven (58%) of these 12 patients: five had portosystemic gradients greater than 12 mm Hg, and two had visible stenosis. In stents with evidence of shunt dysfunction, the average maximal flow velocity was 145 cm/sec (range, 30-351 cm/sec). In stents with no evidence of dysfunction or stenosis, the average maximal velocity was 144 cm/sec (range, 50-190 cm/sec). The differences in these mean velocities were not statistically significant (p = 0.95). Moreover, there was no significant difference between the stenosis and no stenosis groups based on response to changing patient position or respiratory state (p = 0.97). The mean decline in velocities in patients with TIPS dysfunction was 20 cm/sec after sustained deep inspiration, and the mean decline in velocities in patients without shunt dysfunction was 21 cm/sec after sustained deep inspiration.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
It is our experience that obtaining accurate angle-corrected Doppler velocities in TIPS stents can be challenging in some patients. The patients are often large, with distended abdomens and cirrhotic livers that attenuate sound. The acoustic window to the TIPS stent can be limited, and the orientation of the TIPS stent is sometimes perpendicular to the axis of the sound beam when the acoustic window is optimal. Doppler measurements are often made at substantial depths. These complications can try the inventiveness of sonographers and sonologists, who routinely change the patient's position and respiratory state to obtain angle-corrected velocities within this stent. Indeed, at least one study has advocated the use of different positions to facilitate imaging [4].
The physiologic effects of respiratory state on the measured velocities in the TIPS stents may be explained by alterations in venous resistance produced by fluctuating intraabdominal and intrathoracic pressures. When examined in quiet breathing, the venous flow through the TIPS stent is typically monophasic and continuous, but can be seen to increase slightly with inspiration (decreased intrathoracic pressure) and decrease slightly with expiration (increased intrathoracic pressure). This slow undulation of venous flow is a common finding in systemic veins. With deep and sustained inspiration, however, venous flow though the TIPS drastically decreases (Table 1). We believe this respiratory state is tantamount to Valsalva's maneuver, which increases intrathoracic and intraabdominal pressures and tends to collapse abdominal veins, impeding blood flow from the mesenteric and portal venous systems into the TIPS stent.
Theoretically, one might expect venous flow to be diminished in the upright position relative to the supine or decubitus positions. Hydrostatic pressure results from the weight of the blood in the vessels and depends on the position of the patient [13]. Venous return is decreased when the patient is sitting upright and the right atrium is higher than the abdominal veins. Our study did not, however, show a significant effect for patient position. If there are, in fact, velocity differences in TIPS stents at different patient positions, our results indicate that the magnitude of such differences is small and therefore unlikely to be of clinical importance.
Although it is possible that the velocity change in the TIPS in response to respiration could have diagnostic value, we were unable to show a significant difference in the magnitude of that velocity change between patent shunts and malfunctioning shunts. A larger number of patients undergoing portal venography might be needed to discern such a difference.
In summary, this study shows the important effect of respiratory state on the measurement of velocities in TIPS stents. Patient position, in contrast, did not exert a verifiable effect on the mid TIPS velocity, and even if such an effect were present, our results indicate that the magnitude of the effect is small. These findings may help to explain the variations of normal velocity ranges that have been reported from different institutions. Perhaps laboratories reporting a higher normal range of velocities tended to study their patients in quiet respiration or expiration, whereas laboratories reporting lower ranges tended to study their patients in inspiration. This is just speculation because prior reports do not indicate the respiratory state of the patient during their studies.
It is now clear from our study that the respiratory state of the patient should be stipulated in TIPS study protocols and explicitly documented in the final report of each TIPS examination. We hope these steps will help standardize Doppler protocols for TIPS and allow more reliable velocity comparisons between serial studies on a single patient and between patients at different institutions.
Acknowledgments
We thank Susan Murray for assistance with manuscript preparation and David
H. DeLong for a helpful discussion.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Tchelepi, P. W. Ralls, R. Radin, and E. Grant Sonography of Diffuse Liver Disease J. Ultrasound Med., September 1, 2002; 21(9): 1023 - 1032. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A S Shaw, S M Ryan, R C Beese, S Norris, M Bowles, M Rela, and P S Sidhu Liver transplantation Imaging, August 1, 2002; 14(4): 314 - 328. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. G. Sheiman, T. Vrachliotis, D. P. Brophy, and B. J. Ransil Transmitted Cardiac Pulsations as an Indicator of Transjugular Intrahepatic Portosystemic Shunt Function: Initial Observations Radiology, July 1, 2002; 224(1): 225 - 230. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
