Am J Physiol Gastrointest Liver Physiol 287: G830 –G835, 2004;
10.1152/ajpgi.00015.2004.
Mechanics and hemodynamics of esophageal varices
during peristaltic contraction
Larry S. Miller,1 Joseph K. Kim,1 Qing Dai,1 Jyothi Mekapati,1 James Izanec,1 Chan Chung,1
Ji-Bin Liu,4 Andrew Sanderson,1 Matt Bohning,1 Josh Desipio,1 Jasneet Gandegok,1
Justin J. Harberson,1 Carson Schneck,1 Mark A. Nicosia,3 Vinod Thangada,1 Beje Thomas,1
Brian Copeland,1 Elan Miller,1 Aaron Miller,1 Naji Ahmed,1 and James G. Brasseur2
1
Temple University Hospital, Philadelphia 19140; 2The Pennsylvania State University,
University Park 16802-1412; 4Thomas Jefferson University Hospital, Philadelphia, Pennsylvania 19107;
and 3University of Minnesota, Minneapolis, Minnesota 55455
Submitted 13 January 2004; accepted in final form 15 June 2004
tension is an inwardly directed force opposing an outwardly
directed expanding force. Wall tension can be calculated by the
Laplace equation. The Laplace equation (WT ⫽ pv ⫺ pe ⫻ r/w)
states that the wall tension (WT) is equal to the transmural
pressure difference (pv ⫺ pe) (where pv is the intravariceal
pressure and pe is the esophageal lumen pressure) times the
radius of the varix (r) divided by the wall thickness (w). Thus
this force is directly proportional to the transmural pressure
difference and the radius of the varix and inversely proportional to the wall thickness of the varix.
No methods are presently available that can measure
variceal pressures during a peristaltic contraction of the esophagus (1–3, 7, 12). Our hypothesis, based on the data generated
in this study, states that variceal pressure and wall tension
increase dramatically during esophageal peristalsis as a natural
consequence of the anatomy and physiology of the esophagus
and of the esophageal venous plexus.
The objective of this study was to estimate variceal pressure
and wall tension and to characterize the mechanical and hemodynamic behavior of esophageal varices during peristaltic
contractions. This was done through measurements on patients
with esophageal varices and in a model varix system that
simulates intravariceal pressure generated during peristalsis.
MATERIALS AND METHODS
Patient Studies
occurs when an expanding force
within the varix exceeds the maximum wall tension. Wall
This study was approved by the internal review board at Temple
University Hospital. Nine patients (8 men and 1 woman with a mean
age of 53.1 ⫾ 10.9 yr) with cirrhosis, portal hypertension, and
esophageal varices documented on prior outside endoscopy were
evaluated in this study. The etiology of the cirrhosis was hepatitis C
in two patients, hepatitis B in two patients, alcohol in two patients,
hepatitis B and C in one patient, alcohol and hepatitis C in one patient,
and cryptogenic in one patient. None of the patients was treated with
beta blockers during the study as per the referring physicians. None of
the patients had bled from esophageal varices before the study.
Esophageal varices were imaged in cross section with high-frequency endoluminal sonography, using a 20-MHz ultrasonography
transducer (Olympus, Tokyo, Japan or Microvasive, Boston Scientific, Boston, MA). The transducer produces a real-time 360° crosssectional ultrasound image of the esophagus. Real-time images were
recorded on Super VHS videotape using a Kayelemetrics swallowing
workstation (Kayelemetrics, Lincoln Park, NJ).
Address for reprint requests and other correspondence: L. S. Miller, Dept. of
Gastroenterology, Temple Univ. Hospital, 3401 North Broad St., Philadelphia,
PA 19140 (E-mail: vivi@astro.temple.edu).
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked “advertisement”
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
esophageal varices; simultaneous ultrasound and manometry; variceal
bleeding
ESOPHAGEAL VARICEAL BLEEDING
G830
0193-1857/04 $5.00 Copyright © 2004 the American Physiological Society
http://www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.3 on October 2, 2016
Miller, Larry S., Joseph K. Kim, Qing Dai, Jyothi Mekapati,
James Izanec, Chan Chung, Ji-Bin Liu, Andrew Sanderson, Matt
Bohning, Josh Desipio, Jasneet Gandegok, Justin J. Harberson,
Carson Schneck, Mark A. Nicosia, Vinod Thangada, Beje Thomas,
Brian Copeland, Elan Miller, Aaron Miller, Naji Ahmed, and James
G. Brasseur. Mechanics and hemodynamics of esophageal varices during peristaltic contraction. Am J Physiol Gastrointest Liver Physiol 287:
G830 –G835, 2004; 10.1152/ajpgi.00015.2004.—Our hypothesis states
that variceal pressure and wall tension increase dramatically during
esophageal peristaltic contractions. This increase in pressure and wall
tension is a natural consequence of the anatomy and physiology of the
esophagus and of the esophageal venous plexus. The purpose of this
study was to evaluate variceal hemodynamics during peristaltic contraction. A simultaneous ultrasound probe and manometry catheter was
placed in the distal esophagus in nine patients with esophageal varices.
Simultaneous esophageal luminal pressure and ultrasound images of
varices were recorded during peristaltic contraction. Maximum variceal
cross-sectional area and esophageal luminal pressures at which the varix
flattened, closed, and opened were measured. The esophageal lumen
pressure equals the intravariceal pressure at variceal flattening due to
force balance laws. The mean flattening pressures (40.11 ⫾ 16.77
mmHg) were significantly higher than the mean opening pressures
(11.56 ⫾ 25.56 mmHg) (P ⱕ 0.0001). Flattening pressures ⬎80 mmHg
were generated during peristaltic contractions in 15.5% of the swallows.
Variceal cross-sectional area increased a mean of 41% above baseline
(range 7– 89%, P ⬍ 0.0001) during swallowing. The peak closing
pressures in patients that experience future variceal bleeding were significantly higher than the peak closing pressures in patients that did not
experience variceal bleeding (P ⬍ 0.04). Patients with a mean peak
closing pressure ⬎61 mmHg were more likely to bleed. In this study,
accuracy of predicting future variceal bleeding, based on these criteria,
was 100%. Variceal models were developed, and it was demonstrated
that during peristaltic contraction there was a significant increase in
intravariceal pressure over baseline intravariceal pressure and that the
peak intravariceal pressures were directly proportional to the resistance at
the gastroesophageal junction. In conclusion, esophageal peristalsis in
combination with high resistance to blood flow through the gastroesophageal junction leads to distension of the esophageal varices and an
increase in intravariceal pressure and wall tension.
ESOPHAGEAL VARICES DURING PERISTALTIC CONTRACTION
and the hypoechoic blood within the varix. The readers of the images
were blinded to the pressure data. Pressure measurements were read
directly from the data stored and presented on the swallowing workstation.
Ultrasound imaging of the varices, during swallowing of water,
showed that the varices initially increased in size, then decreased in
size, and then flattened, closed, and opened sequentially. Variceal
closure was defined as the first point, on the ultrasound image during
the peristaltic contraction at which the hypoechoic blood within the
varix was no longer visible in the image. Peak closing pressure was
defined as the peak esophageal lumen pressure at which a particular
varix closed. Variceal flattening was defined as the point at which the
exposed (esophageal lumen) side of the varix flattened during the
peristaltic contraction (Fig. 2). Because the videotape of the varices
can be run in a forward or backward direction, variceal flattening was
usually determined by finding the point of variceal closure and then
backing the videotape up to the point of variceal flattening. Variceal
opening was defined as the first point at which the hypoechoic blood
in the variceal lumen was again visible after variceal closing. The
points at which the varix closed, flattened, and opened were identified
with the corresponding pressures on the esophageal lumen pressure
curves. Variceal flattening, closing, and opening pressures were referenced to baseline esophageal pressure in the resting state. Varices
were labeled on the ultrasound image to identify the same varix during
multiple swallows. The amplitude of the peak esophageal lumen
pressure during the peristaltic contraction was recorded and correlated
with the variceal closing pressure to determine whether a relationship
exists between the force of the peristaltic contraction and the intravariceal pressure.
Fig. 1. Simultaneous ultrasound images and manometry pressure curves. A: display screen on the Kayelemetrics workstation
with the esophagus at rest. The display screen shows the
ultrasound image. Note the large open varix is marked. The
esophageal lumen pressure is 12.4 mmHg. B: display screen
showing simultaneous ultrasound and manometry during the
peristaltic contraction. Note the flattened varix is marked. The
esophageal lumen pressure at exactly the point of variceal
flattening is 94.4 mmHg. Intravariceal pressure and the esophageal lumen pressure are equal at variceal flattening.
AJP-Gastrointest Liver Physiol • VOL
287 • OCTOBER 2004 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.3 on October 2, 2016
A 3F angiography catheter was glued to the ultrasound catheter to
measure pressure in the esophageal lumen. A 1-mm side port was
made in the angiography catheter at the same level as the ultrasound
transducer. The distal end of the catheter was closed with silicon glue,
whereas the proximal end was attached to a water-perfused manometry system (Arndorfer, Milwaukee, WI). Water was perfused at a rate
of 0.5 ml/min at a pressure of 15 lb./in.2. The manometry system was
attached to a Kayelemetrics swallowing workstation, which was then
used to synchronize the pressure tracings with the corresponding
ultrasound images (Fig. 1).
All studies were performed after an overnight fast with subjects in
a supine position and with the patient’s head elevated at 30°. Viscous
lidocaine (1%) was administered for nasal anesthesia. The dual 20
MHz ultrasound/manometry catheter was passed transnasally through
the esophagus and into the proximal stomach. The dual catheter was
then pulled into the esophageal body at a level above the gastroesophageal junction where the varices appeared the largest. The transducer
was stabilized in position at this level, and resting esophageal lumen
pressure (with breathing) was measured for at least 10 s and then used
as the zero baseline. Subjects were asked to repeatedly swallow a
bolus of 5 ml of water, with at least 1 min between swallows, while
the investigators continuously acquired ultrasound images and simultaneous pressure data.
Still images were digitized and analyzed by using Image Pro Plus
software (Media Cybernetics, Silver Spring, MD). Incomplete or
inadequate tracings or images due to artifact were not used for
evaluation. The maximum cross-sectional area of the varix at rest and
at maximum distension was calculated by the Image Pro Plus software, from the outlined still images of the varix. The image of each
varix was outlined at the border of the hyperechoic inner variceal wall
G831
G832
ESOPHAGEAL VARICES DURING PERISTALTIC CONTRACTION
Fig. 2. Ultrasound images of varices. A: ultrasound image
showing a varix fully opened at rest. B: same varix at the point
of variceal flattening during the peristaltic contraction.
Esophageal Variceal Swallowing Models
The varix model (Fig. 3) consisted of a thin narrow latex balloon,
used to simulate a varix, placed inside a thicker and wider polyethylene tube, used to simulate the esophagus. The size, consistency, and
material used for the model varix and model esophagus were chosen
to simulate an actual esophageal varix and an actual esophagus as
closely as possible. The latex balloon was attached to an open
hydrostatic reservoir so that the balloon was filled with water at a
constant pressure. The resting model varix pressure (pressure in the
varix with the varix at rest) was chosen to simulate actual intravariceal
resting pressures. A simultaneous ultrasound probe and manometry
catheter was placed in the distal portion of the model varix and a
pressure transducer was placed in the model esophagus at the same
level 1 cm from a plastic electrical tie. The electrical tie was used to
narrow the balloon to simulate the resistance due to the palisade
vessels at the gastroesophageal junction.
The first varix model was tested at a resting pressure of 16.7 mmHg
in the varix. These experiments were performed to determine whether
the intraluminal pressure in the esophagus and intravariceal pressure
correspond during flattening of the varix. A plastic electrical tie was
placed at the distal end of the varix to restrict the varix by 74% of the
original cross-sectional area of the balloon. A 1-kg weight was slid
over the model esophagus to push a 2-ml gel bolus (a mixture of
ultrasound gel and coffee creamer) at a constant velocity of 2.5 cm/s
toward the ultrasound and pressure transducers to simulate a fluid
bolus swallow and a peristaltic contraction wave. The amplitude and
the configuration of the pressure wave closely resembled an actual
peristaltic pressure wave in patients. In addition, the velocity of the
peristaltic wave was set to simulate the actual propagation velocity of
a peristaltic contraction. Fifteen simulated swallows were performed.
Simultaneous ultrasound images and pressure data during these swallows were recorded on Super VHS videotape using the Kayelemetrics
swallowing workstation. Two blinded readers independently determined when the model varix flattened on the recorded ultrasound
image in exactly the same manner as the flattening pressure was
determined in actual varices. The pressures within the model varix and
model esophageal lumen were recorded at that point.
A second study was performed by using the same varix model to
determine the effects of varying the resistance of the distal portion of
the varix on the pressure and cross-sectional area of the varix during
Fig. 3. A: a varix model constructed of a latex balloon (the
varix) inside a polyethelene tube (the esophagus). Various
restrictions in the distal portion of the balloon were used to
simulate the palisade zone. B: enlargement of A.
AJP-Gastrointest Liver Physiol • VOL
287 • OCTOBER 2004 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.3 on October 2, 2016
The investigators (J. Kim and V. Thangada) measuring the variceal
closing, flattening, and opening pressures on the basis of the ultrasound images were blinded to the patient’s clinical status.
Patients were contacted every 6 mo after the initial study to
determine whether they experienced any bleeding or were hospitalized for bleeding. If the patients were hospitalized for bleeding, their
medical records were reviewed to determine the site of bleeding
(variceal vs. nonvariceal) and the time of bleeding with respect to the
original simultaneous ultrasound/manometry test. Peak closing pressures in patients that bled were compared with peak closing pressures
in patients that did not bleed using an unpaired Student’s t-test.
ESOPHAGEAL VARICES DURING PERISTALTIC CONTRACTION
G833
Fig. 4. Ultrasound images demonstrate an increase in crosssectional area during peristaltic contraction. A: image showing
a varix (arrow) while the esophagus is at rest. B: image showing
the same varix (arrow) during the peristaltic contraction. Note
the increase in cross-sectional area during the peristaltic contraction.
Statistical Analysis
Statistics are reported as means ⫾ SD. Statistical significance is
analyzed by using the paired Student’s t-test and Pearson correlation
coefficient. P ⬍ 0.05 was considered significant.
RESULTS
Clinical Study
Forty varices and sixty-three swallows were evaluated in
nine patients. The mean closing pressure (43.8 ⫾ 23.8) and
flattening pressure (40.1 ⫾ 16.8) were significantly higher than
the mean opening pressure (11.6 ⫾ 25.6) (P ⬍ 0.0001).
Flattening pressures ⬎80 mmHg (relative to resting esophageal
pressure) occurred in varices during peristaltic contractions in
15.5% of swallows. Flattening pressures ⬎100 mmHg relative
to the resting esophageal pressure occurred in 4.5% of swallows. The interobserver variability, between the two blinded
readers, for the closing pressure was very low with a correlation coefficient value of r ⫽ 0.97.
There was a strong correlation between the peak esophageal
lumen pressure and the variceal closing pressure in the seven
varices in which three or more swallows could be evaluated (r
between 0.78 and 0.99). There was no correlation between
variceal closing or flattening pressure and the maximum distension of the varix during the peristaltic contraction (r ⫽
0.10). The varices distended, with a mean peak increase in
cross-sectional surface area of 41% (range 7– 89%), from a
baseline mean of 0.29 cm2 to an average maximum of 0.41 cm2
during the peristaltic contraction (an increase in average
variceal radius of 64%; P ⬍ 0.0001; Fig. 4).
On clinical follow-up, four patients (44%) bled from esophageal varices an average of 11.5 ⫾ 7.55 mo after the ultrasound/manometry test. The patients that bled had a significantly higher mean peak closing pressure (83.15 ⫾ 18.38
mmHg) than the five patients that did not bleed (53.2 ⫾ 7.67
mmHg) after a mean of 16.4 ⫾ 0.89 mo (P ⬍ 0.04). With the
use of 61 mmHg as a cutoff for mean peak closing pressure, the
sensitivity, specificity, and positive and negative predictive
value for predicting future variceal bleeding were all 100%.
Varix Model Results
Varix model to determine the flattening pressures (74% restriction). The mean intravariceal pressure at the point at which
the varix wall flattened was 63.1 ⫾ 15.4 and 63.1 ⫾ 14.9
mmHg (Fig. 5). The mean esophageal pressure at the point at
which the varix wall flattened was 63.4 ⫾ 15.9 and 62.9 ⫾ 14.9
mmHg. The mean percent difference between the actual varix
model pressure and the esophageal lumen pressure at the point
at which the varix wall flattened was 3.8 ⫾ 2.8 and 4.5 ⫾
2.8%. The correlation between the model esophageal and
model varix pressures at the point at which the varix wall
Fig. 5. Ultrasound images of the varix model during the peristaltic contraction. A: ultrasound image showing the varix
model at rest. B: ultrasound image showing the varix model at
the point of flattening during the peristaltic contraction. At the
point of flattening, the intravariceal pressure equals the esophageal lumen pressure.
AJP-Gastrointest Liver Physiol • VOL
287 • OCTOBER 2004 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.3 on October 2, 2016
the simulated peristaltic contraction. The latex (varix) balloon was
attached to the open hydrostatic reservoir so that the balloon was filled
with water at a stable pressure of 10 mmHg. Four restrictions with
electrical ties were used to decrease the cross-sectional area of the
distal portion of the model varix (47, 67, 77, and 87% of the
cross-sectional surface area of the original latex balloon). An individual balloon model was set up for each restriction. Ten simulated
swallows were performed for each restriction. Peak pressures and
peak distensions were correlated to the percent restriction in the varix
model using a Pearson correlation coefficient.
G834
ESOPHAGEAL VARICES DURING PERISTALTIC CONTRACTION
DISCUSSION
The simultaneous ultrasound and manometry device, used in
this study to measure variceal pressure during peristaltic contraction, is based on prior work using high resolution endoluminal ultrasound (4, 5, 8, 9, 11, 13, 14) and simultaneous
ultrasound and manometry to evaluate esophageal motility (6,
10, 15, 16).
Force balance laws explain why, at the point of variceal
flattening, the esophageal lumen pressure is equal to the intra-
Fig. 6. Simultaneous intraluminal pressure tracings of the model varix and
model esophagus. Graph shows time on the x-axis and pressure on the y-axis.
The intravariceal pressure is in black and the esophageal lumen pressure is in
gray. Note the peak intravariceal pressure occurs well before the flattening
pressure (the point at which the lines cross over each other is the point at which
the varix flattens). The maximum transmural pressure difference occurs coincidently with the peak variceal cross-sectional area. Thus the wall tension is
greatest at this point. The mean flattening pressure was found to always be on
the down slope of the variceal pressure curve a mean of 34% less than the
mean peak intravariceal pressure.
AJP-Gastrointest Liver Physiol • VOL
Fig. 7. A: graph of the %restriction at the gastroesophageal junction vs. and the
model mean peak intravariceal pressure. In this graph, variceal restriction is on
the x-axis and peak pressure is on the y-axis. The mean peak pressure generated
in the distal portion of the varix during the peristaltic contraction was directly
proportional to the restriction in cross-sectional surface area at the gastroesophageal junction (r ⫽ 0.98). These pressures were significantly higher than
the baseline pressures (P ⬍ 0.0001) at physiological resistances (77%). B:
graph of the %restriction vs. the peak varix model cross-sectional area. In this
graph %restriction is on the x-axis and peak variceal cross-sectional area is on
the y-axis. The maximum change in peak variceal cross-sectional area is
directly proportional to the percent restriction in the distal varix during the
simulated peristaltic contraction (r ⫽ 0.99).
variceal pressure. By using this method, the pressure within the
varix at variceal flattening can be unambiguously inferred from
the measured esophageal lumen pressure. The flattening pressures of the varices during peristalsis were found to be significantly higher than would be expected on the basis of previous
invasive and noninvasive measurements of intravariceal pressure while the esophagus is at rest.
The variceal cross-sectional surface area was measured
throughout the peristaltic contraction sequence and was found
to increase significantly during the phase of lumen distension
just before the rise in the esophageal lumen pressure.
The above observations may be explained by considering the
anatomy and physiology of esophageal varices. Varices, which
are tubular vascular structures, run longitudinally in the body
of the esophagus. The anatomy of esophageal varices in the
body of the esophagus differs dramatically from the anatomy
of the varices at the gastroesophagaeal junction. The venous
structure of esophageal varices in the body of the esophagus is
truncal. Truncal anatomy consists of a few large columns of
varices. The venous structure of esophageal varices at the
gastroesophageal junction consists of palisade vessels. The
palisade veins consist of numerous tiny vessels. The total
cross-sectional area of the varices in the palisade area is ⬃74%
287 • OCTOBER 2004 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.3 on October 2, 2016
flattened was r ⫽ 0.98 and r ⫽ 0.98. The correlation between
the two readers was r ⫽ 0.99 for both the variceal and
esophageal pressures. The model esophageal and model
variceal pressures at the point at which the varix wall flattened
were statistically identical (P ⫽ 0.98). The point at which the
model variceal pressure equaled the model esophageal pressure
(variceal flattening) was always on the upslope of the esophageal model pressure curve and always on the down slope of
the variceal model pressure curve and was less then the mean
peak variceal pressure by an average of 34% (Fig. 6).
Varix model to determine the peak intravariceal pressures
during simulated peristalsis (47, 67, 77, and 87% restriction).
The mean peak pressure, over baseline resting pressure generated in the distal portion of the varix during the simulated
peristaltic contraction was 5.7 mmHg for the 47% restriction,
79.5 mmHg for the 67% restriction, 97.8 mmHg for the 77%
restriction, and 115.2 mmHg for the 87% restriction. The mean
peak pressure was directly proportional to the change in crosssectional surface area at the electrical tie (r ⫽ 0.98). These
pressures were significantly higher than the baseline pressure,
which was 10 mmHg. The mean peak pressure was significantly greater than the baseline pressure for the 77% restriction
(i.e., approximately physiological resistance) (P ⬍ 0.00001).
As the restriction in the distal varix increased, the maximum
increase in variceal cross-sectional area during the simulated
contraction increased in a linear fashion with respect to the
restriction (r ⫽ 0.99 ) (Fig. 7).
ESOPHAGEAL VARICES DURING PERISTALTIC CONTRACTION
AJP-Gastrointest Liver Physiol • VOL
may have negative consequences in terms of variceal bleeding.
It may be possible, in the future, to use this technique to
measure variceal pressures during peristaltic contraction to
predict future variceal bleeding.
REFERENCES
1. Bosch J, Bordas JM, Rigau J, Viola C, Mastai R, Kravetz D, Navas M,
and Rodes J. Noninvasive measurement of the pressure of esophageal
varices using an endoscopic gauge: comparison with measurements by
variceal puncture in patients undergoing endoscopic sclerotherapy. Hepatology 6: 667– 672, 1986.
2. Gertsch P, Fischer G, Kleber G, Wheatley AM, Geigenberger G, and
Sauerbruch T. Manometry of esophageal varices: comparison of an
endoscopic balloon technique with needle puncture. Gastroenterology
105: 1159 –1166, 1993.
3. Gertsch PH and Meister JJ. Pressure measurement in esophageal varices: preliminary report on a new non-invasive method. Gut 28: 1162–
1165, 1987.
4. Jackson WF, Adrain AL, Black M, and Miller LS. Calculation of
esophageal variceal wall tension by direct sonographic and manometric
measurements. Gastrointest Endosc 50: 247–251, 1999.
5. Liu JB, Miller LS, Feld RI, Barbarevech CA, Needleman L, and
Goldberg BB. Gastric and esophageal varices: 20 MHz transnasal endoluminal ultrasound. Radiology 187: 363–366, 1993.
6. McCray WH Jr, Chung C, Parkman H, and Miller LS. Use of
simultaneous high-resolution endoluminal sonography (HRES) and manometry to characterize high-pressure zone of distal esophagus. Dig Dis
Sci 45: 1660 –1666, 2000.
7. Miller ES, Kim JK, Gondehak J, Hara M, Miller AB, and Miller LS.
A new ultrasound guided esophageal variceal pressure-measuring device.
Am J Gastroenter 99: 1267–1273, 2004.
8. Miller LS. Non-invasive methods and apparatus for in vivo measurement
of intra-variceal pressure and wall tension. U.S. Patent no. 5,878,749,
March 9, 1999.
9. Miller LS, Banson F, Bazir K, Korimilli A, Liu JB, Dewan R, and
Wolfson M, Panganamamula KV, Carrasquillo J, Schwartz J, Chaker
AE, and Black M. Risk of esophageal variceal bleeding based on
endoscopic ultrasound evaluation of the sum of esophageal variceal cross
sectional surface area. Am J Gastroenterol 98: 454 – 459, 2003.
10. Miller LS, Liu JB, Colizzo FP, Han T, Marzano J, Barbareuech C,
Hadwig K, Leung L, and Goldberg BB. Correlation of high-frequency
esophageal ultrasonography and manometry in the study of esophageal
motility. Gastroenterology 109: 832– 837, 1995.
11. Miller LS, Schiano TD, Adrain A, Cassidy M, Liu JB, Ter H, Bellary
SV, Dabezies MA, and Black M. Comparison of high-resolution endoluminal sonography to video endoscopy in the detection and evaluation of
esophageal varices. Hepatology 24: 552–555, 1996.
12. Rigau J, Bosch J, Bordas JM, Navasa M, Mastai R, Kravetz D, Bruix
J, Feu F, and Rodes J. Endoscopic measurement of variceal pressure in
cirrhosis: correlation with portal pressure and variceal hemorrhage. Gastroenterology 96: 873– 880, 1989.
13. Schiano TD, Adrain AL, Cassidy MJ, McCray W, Liu JB, Baranowski RJ, Bellary S, Black M, and Miller LS. Use of high-resolution
endoluminal sonography to measure the radius, and wall thickness of
esophageal varices. Gastrointest Endosc 44: 425– 428, 1996.
14. Schiano TD, Adrain AL, Vega KJ, Liu JB, Black M, and Miller LS.
High-resolution endoluminal sonography assessment of the hematocystic
spots of esophageal varices. Gastrointest Endosc 49: 1424 –1427, 1999.
15. Schiano TD, McCray WH, Liu JB, Baranowski RJ, Bellary SV, Black
M, and Miller LS. In vivo comparison of esophageal varices at and above
the diaphragmatic high pressure zone using high resolution endoluminal
sonography. J Clin Gastroenterol 26: 249 –252, 1998.
16. Wong RF, Bonapace ES Jr, Chung C, Liu JB, Parkman H, and Miller
LS. Simultaneous endoluminal sonography and manometry to assess anal
sphincter complex in normal subjects. Dig Dis Sci 43: 2363–2372, 1998.
287 • OCTOBER 2004 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.3 on October 2, 2016
less than in the truncal zone of the esophagus as determined
previously by Schiano et al. (15).
We suggest that during peristaltic contraction the blood flow
in esophageal varices reverses direction. Blood in varices
normally flows via the gastric varices through the palisade
vessels at the gastroesophageal junction and into the truncal
varices in the body of the esophagus in a caudal to oral
direction (distal to proximal in the esophagus). During the
peristaltic contraction, a wave of pressure is propagated in an
oral to caudal direction (proximal to distal in the esophagus),
pushing the blood from the proximal portion of the varix to the
distal portion of the varix and effectively reversing the blood
flow in the varix. This movement is analogous to fluid being
pushed down the lumen of the esophagus toward the stomach
during swallowing. The resistance to blood flow at the gastroesophageal junction, due to the relative decrease in vessel
cross-sectional area, causes the pressure within the distal truncal varices to rise dramatically during the peristaltic contraction.
The varix model demonstrated excellent correlation between
the flattening pressure as measured by the observers from the
esophageal lumen pressure at the point of variceal flattening
and the actual measured pressure within the model varix. In
addition, there was excellent correlation between the observers.
The model varix study in which the distal resistance was
varied demonstrated that a small change in the resistance at the
distal portion of the varix can have a profound effect on the
pressure generated within the varix during the peristaltic contraction. At a physiological resistance of ⬃77% decrease in
cross-sectional area [based on the Schiano study (15)], the
pressures generated corresponded very closely to the closing
and flattening pressures observed in actual varices.
Another observation demonstrated by the varix model system was that the peak intravariceal pressures during the peristaltic contraction were much higher (an average of 34%
higher) than the variceal flattening pressures (Fig. 6). This
suggests that the flattening pressures measured in the actual
varices probably underestimated the maximum intravariceal
pressures generated during the peristaltic contraction. In addition, the varix model demonstrated that the maximum transmural pressure difference occurred coincidently with the maximum variceal distension during the peristaltic contraction,
thus maximizing the variceal wall tension.
Although the number of patients in this series was small, it
was found that the mean peak closing pressures in patients that
experienced future variceal bleeding were significantly higher
than the mean peak closing pressures in patients that did not
experience variceal bleeding (P ⬍ 0.04). Patients with a mean
peak closing pressure ⬎61 mmHg were found to be more
likely to bleed. The accuracy of predicting future variceal
bleeding in this small series, based on these criteria, was 100%.
In conclusion, taken together, the clinical findings and the
findings of the varix model confirm the hypothesis that peristaltic contraction increases the pressure, cross-sectional area,
and wall tension in the distal portion of esophageal varices and
G835