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


     

Published online before print November 22, 2005, 10.1148/radiol.2381041728
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2381041728v1
238/1/167    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 Google Scholar
Google Scholar
Right arrow Articles by Annet, L.
Right arrow Articles by Van Beers, B. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Annet, L.
Right arrow Articles by Van Beers, B. E.
(Radiology 2005;238:167-175.)
© RSNA, 2005

Gastrointestinal Imaging

Esophageal Varices: Evaluation with Transesophageal MR Imaging—Initial Experience1

Laurence Annet, MD, Frank Peeters, PhD, Yves Horsmans, MD, PhD, Laurent Hermoye, MS, Peter Starkel, MD and Bernard E. Van Beers, MD, PhD

1 From the Diagnostic Radiology Unit and Center for Anatomic, Functional, and Molecular Imaging Research (L.A., F.P., L.H., B.E.V.B.) and the Laboratory of Gastroenterology (Y.H., P.S.), Université Catholique de Louvain, St-Luc University Hospital, Avenue Hippocrate 10, B-1200 Brussels, Belgium. From the 2004 RSNA Annual Meeting. Received October 7, 2004; revision requested December 20; revision received March 18, 2005; accepted April 15. Supported by grant 3.4578.00 from the Fonds National de la Recherche Scientifique, Belgium. Address correspondence to L.A. (e-mail: laurence.annet{at}rdgn.ucl.ac.be).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Purpose: To prospectively use transesophageal magnetic resonance (MR) imaging to determine the morphologic and hemodynamic characteristics of esophageal varices.

Materials and Methods: The study was approved by the ethics committee. All patients gave written informed consent. Forty-two patients (29 men, 13 women; mean age, 58 years ± 11 [standard deviation]) with esophageal varices that were recently demonstrated at endoscopy were included in the study. MR imaging was performed by using a receiver probe that was placed in the esophagus. Black-blood T2-weighted MR images were obtained with cardiac triggering and navigator gating of the right hemidiaphragm. On these images, the maximal diameter, minimal diameter, and surface area of the largest esophageal varix were measured. Periesophageal and paraesophageal varices were recorded. A hemodynamic examination was performed in the last 21 patients to undergo MR imaging, which was used to obtain measurements of flow velocity and rate before and after intravenous injection of 50 µg of octreotide or a placebo. A Kruskal-Wallis test was used to assess differences in the diameter and surface area of the varices according to endoscopic grade. Hemodynamic changes observed after octreotide or placebo injection were compared by using an analysis of variance and a 95% confidence interval.

Results: Periesophageal varices were observed in 36 patients, and paraesophageal varices were observed in 32 patients. The minimal diameter, maximal diameter, and surface area of the esophageal varices at MR imaging differed significantly according to endoscopic grade (P < .001). In the periesophageal varices, the velocity and flow changes caused by octreotide differed significantly from those caused by the placebo (P < .001). A decrease in velocity (mean velocity change, –2.766 cm · sec–1) and flow (mean flow change, –0.455 mL · sec–1) was noted after octreotide injection, but no significant change in velocity (mean velocity change, 0.252 cm · sec–1) or flow (mean flow change, 0.018 mL · sec–1) was noted after placebo injection. The surface area of the varices did not change significantly after octreotide (mean change, –0.771 mm2) or placebo (mean change, –0.015 mm2) injection.

Conclusion: Transesophageal MR imaging is a feasible method to assess the morphologic and hemodynamic characteristics of esophageal varices before and after pharmacologic treatment.

© RSNA, 2005


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Bleeding of the esophageal varices is a major complication of portal hypertension. At least two-thirds of patients with cirrhosis develop esophageal varices, and about 30% of patients with cirrhosis experience variceal bleeding. Despite progress in the treatment of variceal bleeding, the percentage of patients who die during the first hemorrhage and the percentage of deaths that occur after the first hemorrhage owing to rebleeding is 20%–35% (1). Currently, primary and secondary prevention with pharmacologic or endoscopic therapy decreases the risk of variceal hemorrhage (2). Unfortunately, about 40% of patients who are treated with drugs do not respond favorably (3).

Endoscopy is the main procedure that is used to diagnose esophageal varices. Endoscopy can also be used to predict variceal bleeding by showing the diameter of the varices and the presence of red spots. Large periesophageal varices that are related to variceal recurrence after sclerotherapy, however, cannot be assessed with endoscopy (4,5). In addition, portal and collateral blood flow and pressure, which are the main determinants of variceal bleeding and response to treatment, cannot be determined with routine endoscopy.

The efficacy of drug therapy can be assessed by measuring the hepatic venous pressure gradient (6,7). This measurement is invasive because catheterization of the hepatic vein is necessary, and the hepatic venous pressure gradient is not always related to the pressure and flow in the varices themselves (8,9). Therefore, alternative methods to measure flow and pressure in the varices with endoscopic Doppler ultasonography (US) or pressure-sensitive endoscopic gauges have been developed (10–13). Because of large interobserver and interequipment variability, however, these methods are not currently recommended in clinical practice (6).

Transesophageal magnetic resonance (MR) imaging, which was developed to image the thoracic aorta, can also be used to image the mediastinal vessels. The use of a radiofrequency receiver probe in the esophagus near the region of interest improves the signal-to-noise ratio and allows for the evaluation of small vessels (14,15). The aim of our study, therefore, was to prospectively use transesophageal MR imaging to determine the morphologic and hemodynamic characterisitics of esophageal varices.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
Patients
From July 2002 to January 2004, 42 patients (29 men, 13 women; mean age, 58 years ± 11) whose recent (<1 month) endoscopic results demonstrated esophageal varices were included in the study. Exclusion criteria consisted of the usual contraindications to MR imaging (ie, intracerebral aneurysm clips, pacemaker implantation, or severe claustrophobia). The study was approved by the ethics committee of our institution, and all patients gave written informed consent to participate in the study.

The cause of portal hypertension was cirrhosis in 37 patients, regenerative nodular hyperplasia in four patients, and portal stenosis following liver transplantation in one patient. The cause of cirrhosis was alcohol abuse in 23 patients, hepatitis C in seven patients, alcohol abuse and hepatitis C or hepatitis B in three patients, nonalcoholic steatohepatitis in two patients, autoimmune hepatitis in one patient, and primary biliary cirrhosis in one patient. Seven patients who previously underwent band ligation to treat variceal bleeding received ß-blocker medications as a secondary preventive measure. Three patients were treated with ß-blocker medications as a primary preventive measure. The delay between variceal ligation and MR imaging was at least 3 months.

Endoscopy
Endoscopy was performed with a flexible fiberscope (GIFQ160; Olympus-Belgium, Aartselaar, Belgium) by two experienced hepatologists (Y.H. and P.S., with 15 and 7 years of experience, respectively). Prior to endoscopy, patients were given 1–5 mg of midazolam (Dormicum; Roche, Basel, Switzerland). The esophageal varices were graded on a scale of 1 to 3. A grade of 1 indicated that the esophageal varices were flattened during insufflation, a grade of 2 indicated that the esophageal varices were not flattened during insufflation but were separated by areas of normal mucosa, and a grade of 3 indicated that the confluent esophageal varices were not flattened during insufflation (16). A total of 24 patients had grade 1 varices, 14 patients had grade 2 varices, and four patients had grade 3 varices.

MR Imaging
The catheter that was used for the transesophageal probe (Intercept esophageal MR coil; Surgi-Vision, Plymouth, Minn) had a width of 4 F, a length of 75 cm, and an imaging length of 5 cm and was protected by an 8-F sheath. After patients fasted overnight, the transesophageal probe was advanced through the nose to the gastroesophageal junction by a radiologist (L.A.). Lidocaine (Xylocaïne Gel; AstraZeneca, Brussels, Belgium) was used to facilitate sliding of the probe. Proper positioning of the probe was confirmed at fluoroscopy. No medication was administered for probe placement or for MR imaging. No complications were observed during the placement of the esophageal coil or at 1 week after MR imaging.

MR imaging was performed with a 1.5-T imager (Gyroscan Intera; Philips Medical Systems, Best, the Netherlands). A morphologic examination of the lower esophagus was performed first. Scout images were used as a guide to place 10 sections perpendicular to the long axis of the lower esophagus. Black-blood MR images were obtained with a double-inversion fast spin-echo MR imaging sequence (17), which was performed with a section thickness of 3 mm, no intersection gap, an in-plane spatial resolution of 0.57 x 0.57 mm, a flip angle of 90°, a repetition time of two heartbeats, an echo time of 26 msec, an echo train length of 24, echo spacing of 5.4 msec, and a bandwidth of 323 Hz per pixel. Two signals were acquired, and the images were obtained during middiastole. Respiratory gating and tracking were performed with a navigator pulse that was positioned on the right hemidiaphragm. The duration of imaging was approximately 4 minutes depending on the cardiac and respiratory frequency.

A hemodynamic examination was performed in the last 21 patients to undergo MR imaging because we first wanted to assess the feasibility of using transesophageal MR imaging to evaluate the morphologic characteristics of the esophageal varices and to determine the tolerance to the method. Hemodynamic examination was therefore not attempted in the first 21 patients. For hemodynamic examination, patients were randomly assigned to receive either octreotide (Sandostatine; Novartis Pharma, Basel, Switzerland) (n = 11) or a placebo (saline solution) (n = 10). Eight patients in the octreotide group and four patients in the placebo group had grade 1 esophageal varices. The remaining patients in both groups had grade 2 varices. The mean age was similar between the two groups (61 years ± 4 for the octreotide group and 59 years ± 14 for the placebo group) (P > .05). In the octreotide group, two patients received ß-blocker medication as a primary preventative measure and two patients who previously underwent variceal band ligation received ß-blocker medication as a secondary preventative measure.

A section passing through the largest periesophageal varix was selected by one author (L.A.). For this section to be perpendicular to the vessel, care was taken to pass through a segment where the course of the varix was not tortuous. At this level, cine phase-contrast spoiled gradient-echo MR imaging was performed with a section thickness of 3 mm, an in-plane spatial resolution of 0.29 x 0.29 mm, 14/7.8 (repetition time msec/echo time msec), a flip angle of 15°, two signals acquired, a bandwidth of 323 Hz per pixel, and retrospective cardiac gating. Thirty-two to eighty phases were acquired depending on the heart rate. Flow was encoded in the inferior-superior direction by using a velocity-encoding value of 20 cm · sec–1. With this velocity-encoding value, the best velocity resolution in the vessels without aliasing artifacts was obtained. The duration of the sequence was approximately 2 minutes depending on the cardiac frequency. Without moving the patient, we repeated the same sequence immediately after intravenous injection of 50 µg of octreotide, a vasoconstrictor that is known to decrease collateral flow in patients with portal hypertension, or saline solution, which was used as a placebo (18). Heart rate was monitored during the entire examination.

Image Analysis
The minimal and maximal diameters of the largest varix were measured on black-blood MR images. On the same images, the surface of the same varix was measured by manually drawing the contours of the vessel. These measurements were performed by one author (L.A.). The minimal diameters of the largest periesophageal and paraesophageal collateral vessels were also measured. Periesophageal veins were defined as vessels that were located adjacent to the muscularis externa of the esophagus, and paraesophageal veins were defined as vessels that were separated from the muscularis externa of the esophagus (10).

After determining by consensus which periesophageal varix demonstrated the best signal intensity on phase-contrast MR images, two investigators (L.A., B.E.V.B.) independently performed velocity and flow measurements by tracing the regions of interest that encompassed the periesophageal varix on these images. The surface area of the vessel was defined as the surface that comprised this region of interest. Manual adaptation was performed frame to frame to include all moving components of the vessel. Periesophageal rather than esophageal varices were chosen for region of interest placement because these varices had a larger diameter, particularly in patients who had undergone variceal band ligation.

Velocity, as measured in centimeters per second, was obtained on each phase-contrast MR image as the mean pixel value within the region of interest. Mean velocity was calculated as the average of the velocities in all cardiac phases. Flow volume, which is presented in milliliters per second, was calculated by multiplying the cross-sectional area by the mean velocity.

Statistical Analysis
Data are expressed as the mean ± standard deviation. Statistical tests were performed by using a computer software program (SPSS, version 11.0.1; SPSS, Chicago, Ill). A P value of less than .050 was considered to indicate a statistically significant difference. The minimal diameter, maximal diameter, and surface area were measured in the largest varix of each patient and were classified into three groups that corresponded to endoscopic grade. Because the variances of these groups were not equal, a Kruskal-Wallis test rather than an analysis of variance was performed to compare the morphologic parameters between these three groups.

The interobserver variability of the velocity, surface area, and flow measurements was shown graphically, as described by Bland and Altman (19). The within-subject coefficient of variation was calculated as the within-subject standard deviation divided by the mean. The within-subject standard deviation was the square root of the within-subject variance, as calculated by using a one-way analysis of variance (20). This analysis was performed after checking that the standard deviations for an individual patient were unrelated to the mean by calculating the Kendall {tau} rank correlation coefficient. When an individual patient's standard deviations were proportional to the means, the coefficient of variation was obtained after logarithmic transformation. Here, the coefficient of variation was {alpha}{sigma}w – 1, where {alpha}{sigma}w was the antilog of the within-subject standard deviation of the logarithmically transformed data (21).

The changes in surface area and hemodynamic parameters that were caused by the injection of octreotide or placebo were compared by using an analysis of variance. Treatment was used as a fixed factor, and patients and observers were used as random factors. Changes in the hemodynamic parameters that were caused by the treatments were expressed as the mean change (ie, the results obtained after treatment injection minus the results obtained before treatment injection) with a 95% confidence interval (CI).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
Esophageal varices were demonstrated in all patients (Fig 1). The minimal mean diameter of the varices was 1.90 mm ± 1.09, the maximal mean diameter was 2.52 mm ± 1.79, and the mean surface area was 6.02 mm2 ± 9.72. The diameter and surface area of the varices differed significantly between endoscopic grades (Table). Periesophageal varices were observed in 36 patients, and paraesophageal varices were observed in 32 patients. The minimal mean diameter of the largest periesophageal varix was 2.80 mm ± 1.97, and the minimal mean diameter of the largest paraesophageal varix was 4.57 mm ± 3.94.



View larger version (140K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a. Transverse black-blood T2-weighted MR images demonstrate (a) grade 2 and (b) grade 3 esophageal varices (thin arrow) with periesophageal varices (thick arrow). Images were obtained by using a double-inversion fast spin-echo MR imaging sequence (repetition time of two heartbeats, echo time of 26 msec) with respiratory gating and tracking. The central area of high signal intensity is caused by the MR probe.

 


View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b. Transverse black-blood T2-weighted MR images demonstrate (a) grade 2 and (b) grade 3 esophageal varices (thin arrow) with periesophageal varices (thick arrow). Images were obtained by using a double-inversion fast spin-echo MR imaging sequence (repetition time of two heartbeats, echo time of 26 msec) with respiratory gating and tracking. The central area of high signal intensity is caused by the MR probe.

 

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

  		Differences between Morphologic Parameters of Esophageal Varices at MR Imaging

 
The surface area, velocity, and flow in the periesophageal varices could be measured before and after octreotide or placebo injection in the 21 patients who underwent phase-contrast MR imaging (Fig 2). The flow that was observed in these vessels was laminar and pulsatile.



View larger version (104K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a. Transverse (a) phase-contrast MR angiogram and (b) phase-contrast MR image obtained with cine phase-contrast sequence (14/7.8, flip angle of 15°, and retrospective cardiac gating) demonstrate posterior periesophageal varices (thick and thin arrows in a and b). One periesophageal varix (thick arrow in a and b) had reversed flow in the superior-inferior direction on phase-contrast images.

 


View larger version (72K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b. Transverse (a) phase-contrast MR angiogram and (b) phase-contrast MR image obtained with cine phase-contrast sequence (14/7.8, flip angle of 15°, and retrospective cardiac gating) demonstrate posterior periesophageal varices (thick and thin arrows in a and b). One periesophageal varix (thick arrow in a and b) had reversed flow in the superior-inferior direction on phase-contrast images.

 
The Bland and Altman plots (Fig 3), which were used to represent the interobserver variability for measurements of hemodynamic parameters, show that the variability of the surface area and flow measurements increased with the magnitude of the mean measurement. This relationship was confirmed by means of the Kendall {tau} correlation coefficients. These coefficients were –0.110 for velocity (P = .121), 0.276 for surface area (P < .001), and –0.288 for flow (P < .001). Therefore, the coefficients of variation for surface area and flow were calculated after logarithmic transformation. The coefficients of variation were ({surd}0.051)/5.59 = 4% for velocity, antilog {surd}0.003 – 1 = 13% for surface area, and antilog {surd}0.002 – 1 = 12% for flow.



View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a. Bland and Altman scatter plots show interobserver differences for measurements of (a) velocity, (b) surface area, and (c) flow in periesophageal varices. Middle horizontal line represents the mean difference, and upper and lower dashed lines represent the 95% interval of agreement. The mean difference between the two observers is located close to zero for all three parameters. The majority of the differences between the two observers remain within the 95% interval of agreement. Note that for surface area and flow, the differences between the observers is proportional to the average of the measurements.

 


View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b. Bland and Altman scatter plots show interobserver differences for measurements of (a) velocity, (b) surface area, and (c) flow in periesophageal varices. Middle horizontal line represents the mean difference, and upper and lower dashed lines represent the 95% interval of agreement. The mean difference between the two observers is located close to zero for all three parameters. The majority of the differences between the two observers remain within the 95% interval of agreement. Note that for surface area and flow, the differences between the observers is proportional to the average of the measurements.

 


View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3c. Bland and Altman scatter plots show interobserver differences for measurements of (a) velocity, (b) surface area, and (c) flow in periesophageal varices. Middle horizontal line represents the mean difference, and upper and lower dashed lines represent the 95% interval of agreement. The mean difference between the two observers is located close to zero for all three parameters. The majority of the differences between the two observers remain within the 95% interval of agreement. Note that for surface area and flow, the differences between the observers is proportional to the average of the measurements.

 
The measurements obtained by the two observers before and after octreotide or placebo injection are shown in Figure 4. The mean values of the velocity, flow, and surface area before octreotide injection were 8.35 cm · sec–1 ± 4.07, 2.06 mL · sec–1 ± 2.34, and 20.75 mm2 ± 15.41, respectively. The mean values of the velocity, flow, and surface area after octreotide injection were 5.58 cm · sec–1 ± 5.03, 1.61 mL · sec–1 ± 2.40, and 19.98 mm2 ± 17.55, respectively. The mean values of the velocity, flow, and surface area before placebo injection were 3.95 cm · sec–1 ± 2.00, 0.17 mL · sec–1 ± 0.15, and 4.11 mm2 ± 2.51, respectively. The mean values of the velocity, flow, and surface area after placebo injection were 4.20 cm · sec–1 ± 2.34, 0.19 mL · sec–1 ± 0.18, and 4.12 mm2 ± 2.50, respectively.



View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a. Scatter plots show changes in (a, b) velocity, (c, d) flow, and (e, f) surface area for periesophageal varices after octreotide (a, c, e) or placebo (b, d, f) injection. Results are noted for observer 1 ({bullet}) and observer 2 ({circ}). After octreotide injection, a decrease of velocity and flow was observed (a, c). After placebo injection, no decrease in velocity or flow was observed (b, d). No substantial modification in surface area was noted after octreotide or placebo injection.

 


View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b. Scatter plots show changes in (a, b) velocity, (c, d) flow, and (e, f) surface area for periesophageal varices after octreotide (a, c, e) or placebo (b, d, f) injection. Results are noted for observer 1 ({bullet}) and observer 2 ({circ}). After octreotide injection, a decrease of velocity and flow was observed (a, c). After placebo injection, no decrease in velocity or flow was observed (b, d). No substantial modification in surface area was noted after octreotide or placebo injection.

 


View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4c. Scatter plots show changes in (a, b) velocity, (c, d) flow, and (e, f) surface area for periesophageal varices after octreotide (a, c, e) or placebo (b, d, f) injection. Results are noted for observer 1 ({bullet}) and observer 2 ({circ}). After octreotide injection, a decrease of velocity and flow was observed (a, c). After placebo injection, no decrease in velocity or flow was observed (b, d). No substantial modification in surface area was noted after octreotide or placebo injection.

 


View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4d. Scatter plots show changes in (a, b) velocity, (c, d) flow, and (e, f) surface area for periesophageal varices after octreotide (a, c, e) or placebo (b, d, f) injection. Results are noted for observer 1 ({bullet}) and observer 2 ({circ}). After octreotide injection, a decrease of velocity and flow was observed (a, c). After placebo injection, no decrease in velocity or flow was observed (b, d). No substantial modification in surface area was noted after octreotide or placebo injection.

 


View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4e. Scatter plots show changes in (a, b) velocity, (c, d) flow, and (e, f) surface area for periesophageal varices after octreotide (a, c, e) or placebo (b, d, f) injection. Results are noted for observer 1 ({bullet}) and observer 2 ({circ}). After octreotide injection, a decrease of velocity and flow was observed (a, c). After placebo injection, no decrease in velocity or flow was observed (b, d). No substantial modification in surface area was noted after octreotide or placebo injection.

 


View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4f. Scatter plots show changes in (a, b) velocity, (c, d) flow, and (e, f) surface area for periesophageal varices after octreotide (a, c, e) or placebo (b, d, f) injection. Results are noted for observer 1 ({bullet}) and observer 2 ({circ}). After octreotide injection, a decrease of velocity and flow was observed (a, c). After placebo injection, no decrease in velocity or flow was observed (b, d). No substantial modification in surface area was noted after octreotide or placebo injection.

 
The changes in velocity (F = 60.299, P < .001) and flow (F = 27.809, P < .001) that were caused by octreotide injection differed significantly from those that were caused by placebo injection. The surface area changes caused by octreotide injection did not differ significantly from those caused by placebo injection (F = 0.327, P > .05).

After octreotide injection, a decrease in velocity (mean velocity change, –2.766 cm · sec–1; 95% CI: –3.308, –2.224) and flow (mean flow change, –0.455 mL · sec–1; 95% CI: –0.580, –0.330) was observed. After placebo injection, no significant change in velocity (mean velocity change, 0.252 cm · sec–1; 95% CI: –0.317, 0.820) or flow (mean flow change, 0.018 mL · sec–1; 95% CI: –0.113, 0.149) occurred. The surface area of the varices did not change significantly after octreotide injection (mean surface area of change, –0.771 mm2; 95% CI: –2.688, 1.146) or after placebo injection (mean surface area of change, –0.015 mm2; 95% CI: –1.995, 2.026).

One minute after octreotide injection, the heart rate decreased significantly (mean change, –7.24 beats; 95% CI: –10.85, –3.64) and returned to basal values after 3 minutes. No significant modification in heart rate was observed after placebo injection (mean change, –1.33 beats; 95% CI: –5.11, 2.46).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
Patients with portal hypertension and esophageal varices are currently evaluated with esophageal endoscopy and invasive measurements of the hepatic venous gradient. These diagnostic methods have known limitations (22–24). Alternative methods for measurement of the morphologic and hemodynamic characteristics of esophageal varices have been developed. The variceal pressure can be measured directly by puncturing the varix during endoscopy (25). This invasive method cannot be performed routinely. Previously, the transmural pressure of the varices has been measured with pressure-sensitive endoscopic gauges (12). This method is difficult to perform, particularly in patients with small varices, and the accuracy and variability of this technique are limited by vessel size. Therefore, the use of pressure-sensitive gauges has not gained wide acceptance (9,26,27).

Endoscopic US has also been proposed as a way to assess patients with esophageal varices (28,29). This modality has been used mainly to evaluate the dimensions of esophageal and periesophageal varices. The observer dependency of endoscopic US has been reported as a potential limitation (6). Endoscopic Doppler US has also been used in some studies. In most of these studies, only the direction of flow was assessed. Until recently, flow rate has been measured in rare instances only (11,30,31).

Reports on the assessment of the collateral circulation with computed tomography and MR imaging are limited. Few morphologic studies have been performed to assess the presence and dimensions of esophageal varices (32–34).

Blood velocity and flow can be measured in the azygos vein by using MR imaging with external coils (35). The relationship between flow in the azygos vein and that in esophageal varices, however, is debated because the azygos vein also has an inflow coming from the general circulation (36,37).

In our study, the morphologic and hemodynamic characteristics of esophageal varices could be assessed with transesophageal MR imaging before and after pharmacologic treatment. The local MR coil was easy to place and well tolerated in all the patients. No medication was needed to perform the examination. The diameter and surface area of the varices at MR imaging differed significantly according to endoscopic grade. It was also possible to assess periesophageal and paraesophageal collateral vessels, which are prognostic factors of variceal recurrence after sclerotherapy (4,5).

The hemodynamic results that were based on the phase-contrast MR images showed a significant decrease in the velocity and flow of blood in the periesophageal varices after the injection of octreotide. Surface area, however, did not change significantly. The interobserver coefficient of variation was less than 5% for velocity and less than 15% for surface area and flow measurements. This variation can be considered acceptable in a clinical context. Octreotide is an analogue of somatostatin, which is a splanchnic vasoconstrictor that decreases portal and collateral blood flow and portal pressure. Bolus injection of octreotide causes a rapid and intense but transient decrease in portal pressure and azygos blood flow (18). Interindividual variation in the effect of octreotide, however, is expected and has been shown in previous studies that examined the effects of octreotide on variceal pressure (27,38). A transient decrease in heart rate, as observed in our study, is also a known effect of octreotide (18).

The flow that was observed in the periesophageal varices was laminar. No turbulences were observed. This is not unexpected because of the low Reynolds number in small vessels with low flow. However, the flow was also pulsatile, as observed in the azygos vein and the vena cava (35). Therefore, the reported data were obtained by averaging all the values that were acquired during one cardiac cycle. Measurements of velocity and flow in the periesophageal varices were obtained at cine phase-contrast MR imaging by using a non-breath-hold technique rather than a breath-hold technique. It has been previously reported that flow in the azygos vein varies according to variations in intrathoracic pressure, which are related to the respiratory cycle (39). Therefore, flow measurements obtained with the non-breath-hold method, which are a mean of the flow in the different phases of respiration, are more representative of the overall blood flow through the thoracic veins than are the flow measurements in either the full inspiratory or the end expiratory phases.

There are some limitations in this study. First, only patients with esophageal varices were included in the study. Therefore, the sensitivity and specificity of transesophageal MR imaging for the detection of esophageal varices could not be determined. The purpose of our study, however, was to determine the feasibility of using transesophageal MR imaging to assess the grade and hemodynamic characteristics of esophageal varices. Second, the diameter of the largest varix that was observed at MR imaging was compared with the endoscopic grade. The endoscopic grade is a global grade that does not take into account differences in the diameter of individual varices. Therefore, a one-to-one comparison between varices at MR imaging and endoscopy was not feasible. In addition, the interobserver variability of the endoscopic grade is large in some studies (24,40). Even if the endoscopic grade remains an imperfect standard, it has a clinical meaning because a correlation between the endoscopic grade and the risk of variceal bleeding has been reported (41). Third, for ethical reasons, the study was not repeated to assess the intrasubject variability. The absence of significant hemodynamic modification after placebo injection is, however, an indirect indicator of a weak intrasubject variability. Fourth, we did not compare transesophageal MR imaging with another diagnostic method to assess the hemodynamic parameters in the periesophageal varices. As discussed previously, there is no accepted noninvasive method to measure the blood flow parameters in the periesophageal and esophageal varices. In the absence of a reference standard, we compared the hemodynamic modifications before and after octreotide or placebo injection and analyzed the interobserver variability with the Bland and Altman method and the coefficients of variation. The interobserver variability can be explained mainly by the fact that it remains difficult to determine the exact contour of the vessels on phase-contrast MR images. Automated algorithms (42) may be useful to perform this task, but these were not used in the present study.

Our results show that the assessment of the hemodynamic parameters in the periesophageal varices with transesophageal MR imaging is feasible, despite the small size of these veins and their low blood flow. Further studies remain to be performed to validate the use of transesophageal MR imaging in the assessment of blood flow parameters in esophageal varices.

In conclusion, preliminary results on the use of transesophageal MR imaging to assess esophageal varices are encouraging. The coil was easy to place and was well tolerated by the patients. Images with sufficient signal intensity and spatial resolution were obtained to measure the diameter and surface area of the esophageal varices and periesophageal and paraesophageal vessels. It was also feasible to measure short-term hemodynamic changes in periesophageal varices after intravenous injection of a splanchnic vasoconstrictor. We believe that transesophageal MR imaging may have a role in the evaluation of the response to treatment of esophageal varices.


   FOOTNOTES
 

Abbreviations: CI = confidence interval

Author contributions: Guarantors of integrity of entire study, L.A., B.E.V.B.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, L.A., F.P., B.E.V.B.; clinical studies, all authors; statistical analysis, L.A., L.H.; and manuscript editing, L.A., Y.H., B.E.V.B.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Bosch J, Burroughs AK. Clinical manifestations and management of bleeding episodes in cirrhotics. In: Bircher J, Benhamou JP, McIntyre N, Rizzetto M, Rodes J, eds. Oxford textbook of clinical hepatology. 2nd ed. Oxford, England: Oxford University Press, 1999; 671–693.
  2. Lui HF, Stanley AJ, Forrest EH, et al. Primary prophylaxis of variceal hemorrhage: a randomized controlled trial comparing band ligation, propranolol, and isosorbide mononitrate. Gastroenterology 2002;123:735–744.[CrossRef][Medline]
  3. Bureau C, Peron JM, Alric L, et al. "A la carte" treatment of portal hypertension: adapting medical therapy to hemodynamic response for the prevention of bleeding. Hepatology 2002;36:1361–1366.[CrossRef][Medline]
  4. Irisawa A, Saito A, Obara K, et al. Endoscopic recurrence of esophageal varices is associated with the specific EUS abnormalities: severe periesophageal collateral veins and large perforating veins. Gastrointest Endosc 2001;53:77–84.[CrossRef][Medline]
  5. Irisawa A, Obara K, Bhutani MS, et al. Role of para-esophageal collateral veins in patients with portal hypertension based on the results of endoscopic ultrasonography and liver scintigraphy analysis. J Gastroenterol Hepatol 2003;18:309–314.[CrossRef][Medline]
  6. de Franchis R. Updating consensus in portal hypertension: report of the Baveno III Consensus Workshop on definitions, methodology and therapeutic strategies in portal hypertension. J Hepatol 2000;33:846–852.[CrossRef][Medline]
  7. Merkel C, Bolognesi M, Sacerdoti D, et al. The hemodynamic response to medical treatment of portal hypertension as a predictor of clinical effectiveness in the primary prophylaxis of variceal bleeding in cirrhosis. Hepatology 2000;32:930–934.[CrossRef][Medline]
  8. Feu F, Bordas JM, Luca A, et al. Reduction of variceal pressure by propranolol: comparison of the effects on portal pressure and azygos blood flow in patients with cirrhosis. Hepatology 1993;18:1082–1089.[CrossRef][Medline]
  9. Brensing KA, Neubrand M, Textor J, et al. Endoscopic manometry of esophageal varices: evaluation of a balloon technique compared with direct portal pressure measurement. J Hepatol 1998;29:94–102.[CrossRef][Medline]
  10. Irisawa A, Obara K, Sato Y, et al. EUS analysis of collateral veins inside and outside the esophageal wall in portal hypertension. Gastrointest Endosc 1999;50:374–380.[CrossRef][Medline]
  11. Hino S, Kakutani H, Ikeda K, et al. Hemodynamic analysis of esophageal varices using color Doppler endoscopic ultrasonography to predict recurrence after endoscopic treatment. Endoscopy 2001;33:869–872.[CrossRef][Medline]
  12. Bosch J, Bordas JM, Rigau J, et al. Noninvasive measurement of the pressure of esophageal varices using an endoscopic gauge: comparison with measurements by variceal puncture in patients undergoing endoscopic sclerotherapy. Hepatology 1986;6:667–672.[Medline]
  13. Escorsell A, Bordas JM, Feu F, et al. Endoscopic assessment of variceal volume and wall tension in cirrhotic patients: effects of pharmacological therapy. Gastroenterology 1997;113:1640–1646.[CrossRef][Medline]
  14. Shunk KA, Lima JA, Heldman AW, Atalar E. Transesophageal magnetic resonance imaging. Magn Reson Med 1999;41:722–726.[CrossRef][Medline]
  15. Shunk KA, Garot J, Atalar E, Lima JA. Transesophageal magnetic resonance imaging of the aortic arch and descending thoracic aorta in patients with aortic atherosclerosis. J Am Coll Cardiol 2001;37:2031–2035.[Abstract/Free Full Text]
  16. Cales P, Zabotto B, Meskens C, et al. Gastroesophageal endoscopic features in cirrhosis: observer variability, interassociations, and relationship to hepatic dysfunction. Gastroenterology 1990;98:156–162.[Medline]
  17. Simonetti OP, Finn JP, White RD, Laub G, Henry DA. "Black blood" T2-weighted inversion-recovery MR imaging of the heart. Radiology 1996;199:49–57.[Abstract/Free Full Text]
  18. Escorsell A, Bandi JC, Andreu V, et al. Desensitization to the effects of intravenous octreotide in cirrhotic patients with portal hypertension. Gastroenterology 2001;120:161–169.[CrossRef][Medline]
  19. Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res 1999;8:135–160.[Abstract/Free Full Text]
  20. Bland JM, Altman DG. Statistics notes: measurement error. BMJ 1996;313:744.[Free Full Text]
  21. Bland JM, Altman DG. Statistics notes: measurement error proportional to the mean. BMJ 1996;313:106.[Free Full Text]
  22. Keiding S, Vilstrup H. Intrahepatic heterogeneity of hepatic venous pressure gradient in human cirrhosis again [letter]. Scand J Gastroenterol 2002;37:1344.[Medline]
  23. Thalheimer U, Mela M, Patch D, Burroughs AK. Targeting portal pressure measurements: a critical reappraisal. Hepatology 2004;39:286–290.[CrossRef][Medline]
  24. Bendtsen F, Skovgaard LT, Sorensen TI, Matzen P. Agreement among multiple observers on endoscopic diagnosis of esophageal varices before bleeding. Hepatology 1990;11:341–347.[Medline]
  25. Palmer ED. Determination of venous pressure within esophageal varices. J Am Med Assoc 1951;147:570–571.[Abstract/Free Full Text]
  26. Escorsell A, Bordas JM, Castaneda B, et al. Predictive value of the variceal pressure response to continued pharmacological therapy in patients with cirrhosis and portal hypertension. Hepatology 2000;31:1061–1067.[CrossRef][Medline]
  27. Nevens F, Van Steenbergen W, Yap SH, Fevery J. Assessment of variceal pressure by continuous non-invasive endoscopic registration: a placebo controlled evaluation of the effect of terlipressin and octreotide. Gut 1996;38:129–134.[Abstract/Free Full Text]
  28. Miller LS, Schiano TD, Adrain A, et al. Comparison of high-resolution endoluminal sonography to video endoscopy in the detection and evaluation of esophageal varices. Hepatology 1996;24:552–555.[CrossRef][Medline]
  29. Irisawa A, Shibukawa G, Obara K, et al. Collateral vessels around the esophageal wall in patients with portal hypertension: comparison of EUS imaging and microscopic findings at autopsy. Gastrointest Endosc 2002;56:249–253.[Medline]
  30. McCormack TT, Rose JD, Smith PM, Johnson AG. Perforating veins and blood flow in oesophageal varices. Lancet 1983;2:1442–1444.[Medline]
  31. Sukigara M, Ohata M, Komazaki T, Omoto R. Assessment of the effect of respiration on the esophageal variceal blood flow using transesophageal real-time two-dimensional Doppler echography. Hepatology 1988;8:663–667.[Medline]
  32. Matsuo M, Kanematsu M, Kim T, et al. Esophageal varices: diagnosis with gadolinium-enhanced MR imaging of the liver for patients with chronic liver damage. AJR Am J Roentgenol 2003;180:461–466.[Abstract/Free Full Text]
  33. Shimizu T, Namba R, Matsuoka T, et al. Esophageal varices before and after endoscopic variceal ligation: evaluation using helical CT. Eur Radiol 1999;9:1546–1549.[CrossRef][Medline]
  34. Imazu H, Matsui T, Noguchi R, et al. Magnetic resonance angiography for monitoring prophylactic endoscopic treatment of high risk esophageal varices. Endoscopy 2000;32:766–772.[CrossRef][Medline]
  35. Wu MT, Pan HB, Chen C, et al. Azygos blood flow in cirrhosis: measurement with MR imaging and correlation with variceal hemorrhage. Radiology 1996;198:457–462.[Abstract/Free Full Text]
  36. Cales P, Braillon A, Jiron MI, Lebrec D. Superior portosystemic collateral circulation estimated by azygos blood flow in patients with cirrhosis: lack of correlation with oesophageal varices and gastrointestinal bleeding—effect of propranolol. J Hepatol 1985;1:37–46.[CrossRef][Medline]
  37. Kassem AM, Salama ZA, Zakaria MS, Hassaballah M, Hunter MS. Endoscopic ultrasonographic study of the azygos vein before and after endoscopic obliteration of esophagogastric varices by injection sclerotherapy. Endoscopy 2000;32:630–634.[CrossRef][Medline]
  38. Avgerinos A, Armonis A, Rekoumis G, Manolakopoulos S, Argirakis G, Raptis S. The effect of somatostatin and octreotide on intravascular oesophageal variceal pressure in patients with cirrhosis. J Hepatol 1995;22:379–380.[CrossRef][Medline]
  39. Ng WH, Chan YL, Sung JY, Lee YT, Lee SF, Chung SS. Comparison of breath-hold 2D phase-contrast with non breath-hold cine phase-contrast MRA in the assessment of azygos venous blood flow in portal hypertension. MAGMA 2004;16:211–217.[CrossRef][Medline]
  40. Cales P, Pascal JP. Gastroesophageal endoscopic features in cirrhosis: comparison of intracenter and intercenter observer variability. Gastroenterology 1990;99:1189.[Medline]
  41. Rigo GP, Merighi A, Chahin NJ, et al. A prospective study of the ability of three endoscopic classifications to predict hemorrhage from esophageal varices. Gastrointest Endosc 1992;38:425–429.[Medline]
  42. Kozerke S, Botnar R, Oyre S, Scheidegger MB, Pedersen EM, Boesiger P. Automatic vessel segmentation using active contours in cine phase contrast flow measurements. J Magn Reson Imaging 1999;10:41–51.[CrossRef][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2381041728v1
238/1/167    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 Google Scholar
Google Scholar
Right arrow Articles by Annet, L.
Right arrow Articles by Van Beers, B. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Annet, L.
Right arrow Articles by Van Beers, B. E.

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