- Email: [email protected]
Minimally invasive measurement of esophageal variceal pressure and wall tension (with video)
ORIGINAL ARTICLE: Clinical Endoscopy
Minimally invasive measurement of esophageal variceal pressure and wall tension (with video) Anil K. Vegesna, MD, MPH, Chan Y. Chung, MD, Anurag Bajaj, MD, Mansoor I. Tiwana, MD, Ranjitha Rishikesh, MD, Imran Hamid, BS, Amit Kalra, MD, Annapurna Korimilli, MD, Sapna Patel, MD, Rasheed Mamoon, MD, Jahenzeb Riaz, MD, Larry S. Miller, MD Philadelphia, Pennsylvania, USA
Background: There is no simple method to measure intravariceal pressure in patients with esophageal varices. Objective: Our purpose was to develop a new noninvasive technique to measure resting intravariceal pressure and wall tension. Design: A model was developed. A long balloon (varix) was fitted inside an airtight cylinder (esophagus). Fluid ran through the model varices to maintain 5 different constant pressures. An endoscope was placed in the model esophagus, and pressure was increased by air insufflation. The endoscopy and pressure readings from the esophagus and varix were recorded continuously until variceal collapse. Setting: Patient studies were done in an endoscopy suite with the patient under fentanyl and midazolam sedation. Patients: Esophageal pressure was measured during air insufflation in patients with varices until the varices collapsed. EUS was used to measure radius and wall thickness to calculate wall tension. Results: In the varix model, the mean (SD) intraluminal esophageal pressures at variceal flattening for the model varices at 5, 10, 15, 20, and 25 mm Hg were 5.69 (0.34), 11 (0.32), 15.72 (0.51), 21.55 (0.63), and 25.8 (0.14) mm Hg. The correlation between actual and measured variceal pressure in the model at variceal flattening was r Z 0.98. In the patients, a total of 10 varices in 3 patients were evaluated. The mean (SD) for the varices in each subject was 12.16 (2.4), 23.2 (1.3), and 6.5 (2.2) mm Hg for subjects 1, 2, and 3, respectively. Conclusion: Standard endoscopy with air insufflation and manometry can be used as an accurate, simple, and reproducible method to measure intravariceal pressure. (Gastrointest Endosc 2009;70:407-13.)
Copyright ª 2009 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 doi:10.1016/j.gie.2008.11.033
pressures,3 and both of these methods are invasive with the risk of hemorrhage. Earlier minimally invasive methods to measure intravariceal pressure were developed by a number of investigators. One such method used a gauge attached to the tip of an endoscope. This consisted of a hemispheric pressure gauge perfused with nitrogen gas.3 A second method, developed by Gertsch et al,4 used a nonexpanding transparent balloon, which was inflated to collapse the varix wall. A third method, developed by Miller et al,5 used a waterfilled balloon with a US tranducer to determine the flattening pressure. A fourth method developed by Miller et al6 used high-resolution US to measure intravariceal pressure during the peristalsis contraction sequence of the esophagus. All these methods required specialized equipment in addition to standard endoscopy and manometry to perform the variceal pressure measurements. No minimally invasive method to date has been developed to
www.giejournal.org
Volume 70, No. 3 : 2009 GASTROINTESTINAL ENDOSCOPY 407
For many years clinicians have sought a minimally invasive method to measure intravariceal pressure. Measuring intravariceal pressure will allow prediction of variceal bleeding, evaluation of pharmacologic therapy, and monitoring of patients after transjugular intrahepatic portosystemic shunt. The criterion standard for determining variceal pressure has been the needle puncture technique.1,2 Measurement of the portohepatic venous pressure gradient has been used as a surrogate. However, portal pressures do not necessarily reflect intravariceal Abbreviation: HPVG, hepatic portal venous gas. DISCLOSURE: All authors disclosed no financial relationships relevant to this publication.
Esophageal variceal pressure and wall tension
measure intravariceal pressure that does not require special equipment other than an endoscope and a pressure transducer. All the above methods are based on the concept of variceal flattening. When the force (pressure) outside a thinwalled vessel starts to exceed the force inside, the vessel starts to collapse by newtonian force mechanics. Newtonian force balance laws explain why, at the point of initiation of variceal flattening, the esophageal lumen pressure is equal to the intravariceal pressure. Variceal wall tension is defined as an inwardly directed force that opposes an outwardly directed expanding force in the variceal wall. It is thought that rupture of a varix occurs when the expanding force exceeds the vessel’s maximal wall tension. The wall tension of a varix can be calculated by the Laplace equation. Wall tension is directly related to transmural pressure and to the vessel radius and is inversely related to vessel wall thickness. Wall tension is expressed by the following formula: TZtp ðr=wÞ where T is wall tension, tp is transmural pressure, r is radius, and w is wall thickness.
PURPOSE The main purpose of this study was to devise a minimally invasive method to accurately measure intravariceal pressure that is simple in application and does not require specialized equipment other than an endoscope and a manometry transducer. The secondary aim of this study was to measure the variceal wall tension.
METHODS This study consisted of 3 sections. In the first section, a varix model was built to demonstrate the feasibility, accuracy, and reliability of the intravariceal-pressure measuring method. The second section was a small feasibility trial to determine the practicality and ease of use of this technology in patients with esophageal varices. The final section described the use of endoluminal US, in combination with this new variceal pressure measuring method, to calculate the variceal wall tension in patients with esophageal varices.
Variceal model
Vegesna et al
Capsule Summary What is already known on this topic d
d
Esophageal variceal bleeding is associated with poor outcomes. Measurement of intravariceal pressure may predict and help prevent bleeding episodes.
What this study adds to our knowledge d
In a varix model and in 3 patients, endoscopy with air insufflation and manometry provided an accurate, simple, and reproducible method for measuring intravariceal pressure.
Pressure within the model esophagus was measured directly with a pressure transducer in the esophageal model. Both manometry catheters were connected to an Arndorfer perfusion pump set at a perfusion rate of 0.5 mL/min. The pressure signal from the manometry catheters and the video signal from the endoscope were recorded simultaneously on a Kay Elemetrics (Lincoln Park, NJ) swallowing workstation. Air was insufflated into the esophagus through the air-water channel on the endoscope to increase the esophageal pressure. Air insufflation was continued until the increased pressure within the esophagus caused the varices to collapse on endoscopic visualization. A videotape of the endoscopy and pressure readings from the model esophagus were simultaneously recorded (Fig. 2). The method that the investigators used to determine the initiation of variceal flattening was to identify complete collapse on the endoscopic videotape and then to rewind the videotape to the point of the initiation of variceal flattening (Fig. 3). Pressure was then determined in a blinded fashion from the simultaneous manometry readings. Endoscopic variceal flattening (intravariceal pressure) was read from the endoscopic recordings by 2 independent investigators blinded to the pressure. The 2 readers were extensively trained by use of actual videotapes of varices during variceal flattening, and specific criteria were used to determine the exact point in time at which variceal flattening occurred, with use of both the variceal model and actual varices. A Pearson correlation coefficient was used to determine the correlation between the model esophageal lumen pressure and the model varix lumen pressure at variceal flattening. A Pearson correlation coefficient was also used to determine agreement between observers. An average of 25 readings was taken for each of the 5 intravariceal pressures.
An esophageal varix model was developed by using a long latex balloon (varix model) fitted inside an airtight cylinder (model esophagus). Fluid ran through the varix model to maintain constant pressures at 5, 10, 15, 20, and 25 mm Hg. The pressure was generated by placing a water reservoir at various heights above the model with the reservoir open to the atmosphere (Fig.1). Pressure within the model varix was measured by placing a manometry catheter directly within the latex balloon.
The above method was tested in 3 patients (aged 59, 57, and 52 years, all male, etiology of cirrhosis for 2 patients was hepatitis C virus, and 1 patient had alcoholic cirrhosis) with portal hypertension on a total of 10 varices
408 GASTROINTESTINAL ENDOSCOPY Volume 70, No. 3 : 2009
www.giejournal.org
Intravariceal pressure measurement in vivo
Vegesna et al
Esophageal variceal pressure and wall tension
Figure 2. Model of esophagus and varix along with Arndorfer pump and Kay Elemetrics workstation for recording video and manometry pressures.
Figure 1. A, Model esophagus with varix along with endoscope and manometry catheter. B, Schematic presentation of model esophagus with varix, along with endoscope and manometry catheter.
that had never undergone endoscopic treatment. The patients were sedated with midazolam and fentanyl and underwent topical anesthesia with benzocaine. All variceal pressures were measured while the esophagus was at rest. We call this the resting variceal pressure. The endoscope (model 180, Olympus, Center Valley, Pa) was placed in the stomach, and the air in the stomach was removed. A water-perfused manometry catheter attached to an Arndorfer pump was placed through the biopsy channel of the endoscope, and the baseline intragastric pressure was measured. The endoscopic images and the manometry tracing were recorded simultaneously on a Kay Elemetrics swallowing workstation. The endoscope was then withdrawn into the esophagus. Varices in the distal esophagus were localized, and the endoscope was placed in the distal esophagus at the level of the distal portion of the esophageal varices where they appeared largest. Air was insufflated into the esophagus through the air channel of the endoscope to increase the esophageal pressure. The endoscopic processor was set to the lowest air volume. If the pressure was insufficient to flatten the varices, the pressure was increased in a graded manner from medium to high. Air insufflation was continued until the increased pressure in the esophagus caused the varices to collapse on endoscopic visualization. Air leaking into the stomach through the lower esophageal sphincter did not present a problem because the pressure gradually increased in the esophagus despite any air leak. www.giejournal.org
After the endoscopic examination was performed, endoscopic variceal flattening (intravariceal pressure) was determined from the endoscopic recordings by 2 independent investigators blinded to the pressure. The method that the investigators used to determine the initiation of variceal flattening was to identify complete collapse on the endoscopic videotape and then to rewind the tape to the point of the initiation of variceal flattening (Fig. 4). Pressure was then determined from the simultaneous esophageal lumen manometry readings. These pressures were measured 3 times for each varix, and the mean SD of the pressures was calculated and recorded as the intravariceal pressure. A Pearson correlation coefficient was also used to determine the agreement between observers.
Variceal wall tension measurements During the endoscopic procedure, a 20-MHz high-resolution US probe was used to image the distal portion of the esophageal varices. The probe was passed through the biopsy channel of the endoscope after the manometry catheter was removed. We did not measure size or the shape of the varices endoscopically. We measured the variceal circumference sonographically and calculated the variceal radius when the esophagus was at rest (resting variceal circumference and variceal radius). Measurement of the sonographic cross-sectional areas and the circumference of esophageal varices at rest are now considered the criterion standard for ‘‘size’’ measurements of esophageal varices.7,8 To identify the varices on the US image with the varices on the endoscopic image, the picturein-picture feature on the endoscope was used. Measurements of the radius and wall thickness of 10 varices in 3 patients with portal hypertension were performed on Image Pro Plus software (version 6.0, Media Cybernetics, Bethesda, Md) after the images were digitized. Volume 70, No. 3 : 2009 GASTROINTESTINAL ENDOSCOPY 409
Esophageal variceal pressure and wall tension
Vegesna et al
Figure 4. Endoscopic view of varices in vivo during air insufflation showing pre (A) and post (B) variceal flattening. Note that the varices are bulging into the lumen of the esophagus before air insufflation (A) and that the varices are collapsed after air insufflation (B).
RESULTS Model varix results
Figure 3. Endoscopic view of the varix model showing variceal collapse during air insufflation.
Two investigators blinded to each other’s readings and to the intravariceal pressures measured the radius and wall thickness of the varices. Wall thickness and variceal radius were measured from the high-resolution US images with Image Pro Plus software (Fig. 5). Variceal wall tension was calculated with the Laplace equation (Wall tension Z Change in pressure Radius of varix/Wall thickness). Variceal flattening pressure (intravariceal pressure) was measured by the air insufflation method described above for calculation of wall tension. 410 GASTROINTESTINAL ENDOSCOPY Volume 70, No. 3 : 2009
The mean (SD) intraluminal esophageal pressures at the initiation of model variceal flattening for the model varix at 5, 10, 15, 20, and 25 mm Hg pressure were 5.69 (0.34), 11 (0.32), 15.72 (0.51), 21.55 (0.63), and 25.8 (0.14) mm Hg, respectively. The correlation coefficient between the actual model variceal pressure and the measured pressure in the model esophagus at model variceal flattening was r Z 0.98. The correlation coefficient between the 2 readers was r Z 0.99. The mean 2 standard errors for both readers can be seen in Figure 6.
In vivo variceal pressure measurement A total of 10 varices in 3 individual patients were evaluated. The individual varices in each patient flattened at different pressures, indicating that each varix has a different intravariceal pressure associated with it. The mean (SD) for the varices in each subject was 12.16 (2.4), 23.2 (1.3), and 6.5 (2.2) mm Hg for subjects 1, 2, and 3, respectively. The interobserver variability between readers was r Z 0.99.
Wall tension The intravariceal pressure correlated well with the intravariceal radius (r Z 0.85), but not with wall thickness (r Z 0.35). The mean (SD) for the wall tension in subjects www.giejournal.org
Vegesna et al
Figure 5. High-resolution US image of the varices with picture-in-picture endoscopic view. The picture also shows the outlines of the varix circumference and the wall thickness. Note that the inner circle (T1) outlines the inner circumference of the varix. The outer ark (T2) outlines the outer circumference of the varix. The wall thickness and variceal radius can be derived from these measurements.
Esophageal variceal pressure and wall tension
Figure 6. Correlation between the esophageal varix pressures and measured esophageal pressures at the initiation of variceal flattening in the varix model.
Given the poor outcome associated with variceal bleeding, the identification of varices at high risk and the prevention of first bleeding episodes are critical objectives. Screening endoscopy is generally recommended for patients with cirrhosis to determine whether large esophageal varices are present. Currently, a combination of clinical and endoscopic findings, including an advanced Child-Pugh class of cirrhosis, large varices on endoscopy, and the presence of red wale markings on endoscopy, correlate with the risk of a first bleeding episode in patients with cirrhosis.9 Our group has also demonstrated that the total cross-sectional surface area of the varices10 and the pressure generated during peristaltic contraction can be used to predict future variceal bleeding.6 At least 4 minimally invasive methods of measuring intravariceal pressure have been developed that exploit the concept of variceal flattening.2-6 However, all of these methods require additional equipment that is not readily available in most endoscopy units. Air insufflation to flatten varices is a standard procedure performed by almost every endoscopist. In fact, most grading systems of esophageal varices use variceal flattening to define grade 1 varices. This is a rough grading system, given that a high enough pressure will flatten any untreated esophageal varix and that the air pressure or volume on most endoscopes can be adjusted (low, medium, and high). In this study, we used the concept of variceal flattening and quantified the flattening pressure by measuring the esophageal
lumen pressure at the initiation of variceal flattening. We first demonstrated this concept in a model varix system and found the results to be accurate and reliable. We then demonstrated the feasibility of performing this variceal pressure measuring method in patients with virgin esophageal varices. Finally, we combined this variceal pressure measuring method with measurements on high-resolution endoluminal US to calculate the variceal wall tension with the Laplace equation. Force balance laws explain why, at the point of variceal flattening, the esophageal lumen pressure is equal to the intravariceal pressure. With this new method, the pressure within the varix at variceal flattening can be unambiguously inferred from the measured esophageal lumen pressure. The variceal flattening pressures of the model esophagus were found to strongly correlate with the flattening pressures within the model varices during air insufflation. It is clear from the literature that portal pressures do not reflect or correlate with esophageal variceal pressures.3 In fact, one of the reasons a variceal pressure measuring device is needed is because we cannot extrapolate hepatic portal venous gas (HPVG) to variceal pressures directly. We believe that the pressure head within the portal vein is damped by the high-resistance venous circulation (the palisade vessels) when the blood has to pass through these vessels to get from the stomach to the esophagus. This was demonstrated in 2 previous studies.6,11 Therefore, a pressure reading of 6 mm Hg in the esophageal varices is consistent with an HPVG of 12 or higher, which is required for varices to form because the pressure force generated in the portal vein is dissipated in palisade esophageal varices. We cannot validate variceal pressure measurements in vivo by measuring the HPVG because HPVG does not
www.giejournal.org
Volume 70, No. 3 : 2009 GASTROINTESTINAL ENDOSCOPY 411
1, 2, and 3 was 57.86 (9.22), 156.37 (46.79), and 8.29 (2.61) mm Hg, respectively.
CONCLUSIONS
Esophageal variceal pressure and wall tension
Vegesna et al
correlate with the variceal pressure. The only way of currently measuring variceal pressure directly is with a needle puncture method or another noninvasive method that has been previously validated. In this day and age of band ligation, it is no longer considered acceptable or ethical to perform variceal needle puncture to validate a noninvasive method of measuring intravariceal pressure. Luckily, there are other methods that use exactly the same principle (variceal flattening). The accuracy of these methods had been confirmed and validated in the past by use of variceal needle puncture.2 We can therefore extrapolate these findings to our new method and, with a great deal of confidence, state that our method is at least as valid as these other methods. The equalization pressure measured at the time of initial variceal flattening will not change, regardless of leaking air through the gastroesophageal junction or the upper esophageal sphincter, and is independent of the flexibility and distensibility of the esophagus. By newtonian physics and force balance equations, at the initiation of variceal flattening, the pressure inside the varix equals the pressure outside the varix. This has been demonstrated by use of other noninvasive methods for measuring esophageal variceal pressure.2 In fact, the entire concept of using air for measuring intravariceal pressure is based on previously validated studies that used similar methods (variceal flattening).2-6 Although there is often movement in the esophagus, which can make pressure measurement more difficult, the actual flattening pressure does not change on the basis of the flexibility, distensibility, or movement of the esophagus. Because the variceal flattening pressure was measured with the esophagus at rest (resting esophageal pressure), at 1 instant in time by increasing the pressure within the esophageal lumen to above the closing pressure within the perforating veins, we essentially have negated the effect of flow within the perforating vessels at the time of variceal flattening. These small thin-walled vessels close under increased esophageal pressure at a much lower closing pressure than did the large varices that were measured. In addition, by use of this new method it is easy to repeat the pressure measurements over and over again. We determined variceal wall tension in vivo by measuring variceal radius and wall thickness with high-resolution EUS on the basis of the previous work of Schiano et al7 and found a strong correlation between wall tension and variceal radius, but not with wall thickness. These same findings were shown previously by Jackson et al12 with the needle puncture method to measure variceal pressure and high-resolution EUS to measure the radius and wall thickness of the varices. The lack of correlation between wall thickness and variceal size in this study implies that variceal wall tension cannot be accurately estimated by the endoscopic observation of variceal size alone or even by size measurements combined with intravariceal pressure measurements.
We believe that the measurement of the resting variceal pressure has potential utility in evaluating the effectiveness of pharmaceutical therapy in individual patients and in evaluating potential medications that can lower variceal pressure in research studies. In fact, prior investigators have shown a decrease in the resting variceal pressure with b-blocker therapy, thus demonstrating the effectiveness of b-blockers.13 In summary, given the poor outcome associated with variceal bleeding, the identification of varices at high risk and the prevention of first bleeding episodes are critical objectives. Standard endoscopy with air insufflation and manometry can be used as an accurate, simple, and reproducible method to measure intravariceal pressure in a varix model system. This method was tested in a clinical setting to determine the feasibility of measuring intravariceal pressure. Finally, a noninvasive method of calculating variceal wall tension was developed and tested. This new technique may help provide a more accurate method for determining the risk of initial variceal bleeding and for monitoring pharmacologic therapy. We believe that the potentially wide applicability and ease of use make this method of high interest to clinicians and investigators.
412 GASTROINTESTINAL ENDOSCOPY Volume 70, No. 3 : 2009
www.giejournal.org
REFERENCES 1. Kleber G, Sauerbruch T, Fischer G, et al. Pressure of intraesophageal varices assessed by fine needle puncture: its relation to endoscopic signs and severity of liver disease in patients with cirrhosis. Gut 1999;30:228-32. 2. Bosch J, Bordas JM, Rigau J, et al. Non-invasive measurement of the pressure of esophageal varices using an endoscopic gauge: comparison with measurements by variceal puncture in patients undergoing endoscopic sclerotherapy. Hepatology 1986;6:667-72. 3. Rigau J, Bosch J, Bordas JM, et al. Endoscopic measurement of variceal pressure in cirrhosis: correlation with portal pressure and and variceal hemorrhage. Gastroenterology 1989;96:873-80. 4. Gertsch P, Wheatley AM, Maibach R, et al. Experimental evaluation of an endoscopic balloon for manometry of esophageal varices. Gastroenterology 1991;101:1692-700. 5. Miller LS, Dai Q, Thomas A, et al. A new ultrasound-guided esophageal variceal pressure measuring device. Am J Gastroenterol 2004;99:1267-73. 6. Miller LS, Kim JK, Dai Q, et al. Mechanics and hemodynamics of esophageal varices during peristaltic contraction. Am J Physiol Gastrointest Liver Physiol 2004;287:G830-5. 7. Schiano TD, Adrain AL, Cassidy MJ, et al. Use of high-resolution endoluminal sonography to measure the radius and wall thickness of esophageal varices. Gastrointest Endosc 1996;44:425-8. 8. 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;24:552-5. 9. North Italian Endoscopic Club for the Study and Treatment of Esophageal Varices. Prediction of first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices: a prospective multicenter study. N Engl J Med 1988;319:983-9. 10. Miller L, Banson FL, Bazir K, et al. Risk of esophageal variceal bleeding based on endoscopic ultrasound evaluation of the sum of esophageal variceal cross-sectional surface area. Am J Gastroenterol 2003;98:454-9. 11. Schiano DT, McCray HW, Liu J-B, et al. In vivo comparison of esophageal varices at and above the diaphragmatic high pressure zone using high resolution endoluminal sonography. J Clin Gastroenterol 1998;26:249-52.
Vegesna et al 12. Jackson WF, Adrain AL, Black M, et al. Calculation of esophageal variceal wall tension by direct sonographic and manometric measurements. Gastrointest Endosc 1999;50:247-51. 13. Nevens F, Sprengers D, Feu F, et al. Measurement of variceal pressure with an endoscopic pressure sensitive gauge: validation and effect of propranolol therapy in chronic conditions. J Hepatol 1996;24: 66-73.
www.giejournal.org
Esophageal variceal pressure and wall tension
Received June 30, 2008. Accepted November 12, 2008. Current affiliations: Temple University Hospital, Philadelphia, Pennsylvania, USA. Reprint requests: Larry S. Miller, MD, Section of Gastroenterology, 8th Floor, Parkinson Pavilion, Temple University Hospital, Philadelphia, PA 19140.
Volume 70, No. 3 : 2009 GASTROINTESTINAL ENDOSCOPY 413
Minimally invasive measurement of esophageal variceal pressure and wall tension (with video) Anil K. Vegesna, MD, MPH, Chan Y. Chung, MD, Anurag Bajaj, MD, Mansoor I. Tiwana, MD, Ranjitha Rishikesh, MD, Imran Hamid, BS, Amit Kalra, MD, Annapurna Korimilli, MD, Sapna Patel, MD, Rasheed Mamoon, MD, Jahenzeb Riaz, MD, Larry S. Miller, MD Philadelphia, Pennsylvania, USA
Background: There is no simple method to measure intravariceal pressure in patients with esophageal varices. Objective: Our purpose was to develop a new noninvasive technique to measure resting intravariceal pressure and wall tension. Design: A model was developed. A long balloon (varix) was fitted inside an airtight cylinder (esophagus). Fluid ran through the model varices to maintain 5 different constant pressures. An endoscope was placed in the model esophagus, and pressure was increased by air insufflation. The endoscopy and pressure readings from the esophagus and varix were recorded continuously until variceal collapse. Setting: Patient studies were done in an endoscopy suite with the patient under fentanyl and midazolam sedation. Patients: Esophageal pressure was measured during air insufflation in patients with varices until the varices collapsed. EUS was used to measure radius and wall thickness to calculate wall tension. Results: In the varix model, the mean (SD) intraluminal esophageal pressures at variceal flattening for the model varices at 5, 10, 15, 20, and 25 mm Hg were 5.69 (0.34), 11 (0.32), 15.72 (0.51), 21.55 (0.63), and 25.8 (0.14) mm Hg. The correlation between actual and measured variceal pressure in the model at variceal flattening was r Z 0.98. In the patients, a total of 10 varices in 3 patients were evaluated. The mean (SD) for the varices in each subject was 12.16 (2.4), 23.2 (1.3), and 6.5 (2.2) mm Hg for subjects 1, 2, and 3, respectively. Conclusion: Standard endoscopy with air insufflation and manometry can be used as an accurate, simple, and reproducible method to measure intravariceal pressure. (Gastrointest Endosc 2009;70:407-13.)
Copyright ª 2009 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 doi:10.1016/j.gie.2008.11.033
pressures,3 and both of these methods are invasive with the risk of hemorrhage. Earlier minimally invasive methods to measure intravariceal pressure were developed by a number of investigators. One such method used a gauge attached to the tip of an endoscope. This consisted of a hemispheric pressure gauge perfused with nitrogen gas.3 A second method, developed by Gertsch et al,4 used a nonexpanding transparent balloon, which was inflated to collapse the varix wall. A third method, developed by Miller et al,5 used a waterfilled balloon with a US tranducer to determine the flattening pressure. A fourth method developed by Miller et al6 used high-resolution US to measure intravariceal pressure during the peristalsis contraction sequence of the esophagus. All these methods required specialized equipment in addition to standard endoscopy and manometry to perform the variceal pressure measurements. No minimally invasive method to date has been developed to
www.giejournal.org
Volume 70, No. 3 : 2009 GASTROINTESTINAL ENDOSCOPY 407
For many years clinicians have sought a minimally invasive method to measure intravariceal pressure. Measuring intravariceal pressure will allow prediction of variceal bleeding, evaluation of pharmacologic therapy, and monitoring of patients after transjugular intrahepatic portosystemic shunt. The criterion standard for determining variceal pressure has been the needle puncture technique.1,2 Measurement of the portohepatic venous pressure gradient has been used as a surrogate. However, portal pressures do not necessarily reflect intravariceal Abbreviation: HPVG, hepatic portal venous gas. DISCLOSURE: All authors disclosed no financial relationships relevant to this publication.
Esophageal variceal pressure and wall tension
measure intravariceal pressure that does not require special equipment other than an endoscope and a pressure transducer. All the above methods are based on the concept of variceal flattening. When the force (pressure) outside a thinwalled vessel starts to exceed the force inside, the vessel starts to collapse by newtonian force mechanics. Newtonian force balance laws explain why, at the point of initiation of variceal flattening, the esophageal lumen pressure is equal to the intravariceal pressure. Variceal wall tension is defined as an inwardly directed force that opposes an outwardly directed expanding force in the variceal wall. It is thought that rupture of a varix occurs when the expanding force exceeds the vessel’s maximal wall tension. The wall tension of a varix can be calculated by the Laplace equation. Wall tension is directly related to transmural pressure and to the vessel radius and is inversely related to vessel wall thickness. Wall tension is expressed by the following formula: TZtp ðr=wÞ where T is wall tension, tp is transmural pressure, r is radius, and w is wall thickness.
PURPOSE The main purpose of this study was to devise a minimally invasive method to accurately measure intravariceal pressure that is simple in application and does not require specialized equipment other than an endoscope and a manometry transducer. The secondary aim of this study was to measure the variceal wall tension.
METHODS This study consisted of 3 sections. In the first section, a varix model was built to demonstrate the feasibility, accuracy, and reliability of the intravariceal-pressure measuring method. The second section was a small feasibility trial to determine the practicality and ease of use of this technology in patients with esophageal varices. The final section described the use of endoluminal US, in combination with this new variceal pressure measuring method, to calculate the variceal wall tension in patients with esophageal varices.
Variceal model
Vegesna et al
Capsule Summary What is already known on this topic d
d
Esophageal variceal bleeding is associated with poor outcomes. Measurement of intravariceal pressure may predict and help prevent bleeding episodes.
What this study adds to our knowledge d
In a varix model and in 3 patients, endoscopy with air insufflation and manometry provided an accurate, simple, and reproducible method for measuring intravariceal pressure.
Pressure within the model esophagus was measured directly with a pressure transducer in the esophageal model. Both manometry catheters were connected to an Arndorfer perfusion pump set at a perfusion rate of 0.5 mL/min. The pressure signal from the manometry catheters and the video signal from the endoscope were recorded simultaneously on a Kay Elemetrics (Lincoln Park, NJ) swallowing workstation. Air was insufflated into the esophagus through the air-water channel on the endoscope to increase the esophageal pressure. Air insufflation was continued until the increased pressure within the esophagus caused the varices to collapse on endoscopic visualization. A videotape of the endoscopy and pressure readings from the model esophagus were simultaneously recorded (Fig. 2). The method that the investigators used to determine the initiation of variceal flattening was to identify complete collapse on the endoscopic videotape and then to rewind the videotape to the point of the initiation of variceal flattening (Fig. 3). Pressure was then determined in a blinded fashion from the simultaneous manometry readings. Endoscopic variceal flattening (intravariceal pressure) was read from the endoscopic recordings by 2 independent investigators blinded to the pressure. The 2 readers were extensively trained by use of actual videotapes of varices during variceal flattening, and specific criteria were used to determine the exact point in time at which variceal flattening occurred, with use of both the variceal model and actual varices. A Pearson correlation coefficient was used to determine the correlation between the model esophageal lumen pressure and the model varix lumen pressure at variceal flattening. A Pearson correlation coefficient was also used to determine agreement between observers. An average of 25 readings was taken for each of the 5 intravariceal pressures.
An esophageal varix model was developed by using a long latex balloon (varix model) fitted inside an airtight cylinder (model esophagus). Fluid ran through the varix model to maintain constant pressures at 5, 10, 15, 20, and 25 mm Hg. The pressure was generated by placing a water reservoir at various heights above the model with the reservoir open to the atmosphere (Fig.1). Pressure within the model varix was measured by placing a manometry catheter directly within the latex balloon.
The above method was tested in 3 patients (aged 59, 57, and 52 years, all male, etiology of cirrhosis for 2 patients was hepatitis C virus, and 1 patient had alcoholic cirrhosis) with portal hypertension on a total of 10 varices
408 GASTROINTESTINAL ENDOSCOPY Volume 70, No. 3 : 2009
www.giejournal.org
Intravariceal pressure measurement in vivo
Vegesna et al
Esophageal variceal pressure and wall tension
Figure 2. Model of esophagus and varix along with Arndorfer pump and Kay Elemetrics workstation for recording video and manometry pressures.
Figure 1. A, Model esophagus with varix along with endoscope and manometry catheter. B, Schematic presentation of model esophagus with varix, along with endoscope and manometry catheter.
that had never undergone endoscopic treatment. The patients were sedated with midazolam and fentanyl and underwent topical anesthesia with benzocaine. All variceal pressures were measured while the esophagus was at rest. We call this the resting variceal pressure. The endoscope (model 180, Olympus, Center Valley, Pa) was placed in the stomach, and the air in the stomach was removed. A water-perfused manometry catheter attached to an Arndorfer pump was placed through the biopsy channel of the endoscope, and the baseline intragastric pressure was measured. The endoscopic images and the manometry tracing were recorded simultaneously on a Kay Elemetrics swallowing workstation. The endoscope was then withdrawn into the esophagus. Varices in the distal esophagus were localized, and the endoscope was placed in the distal esophagus at the level of the distal portion of the esophageal varices where they appeared largest. Air was insufflated into the esophagus through the air channel of the endoscope to increase the esophageal pressure. The endoscopic processor was set to the lowest air volume. If the pressure was insufficient to flatten the varices, the pressure was increased in a graded manner from medium to high. Air insufflation was continued until the increased pressure in the esophagus caused the varices to collapse on endoscopic visualization. Air leaking into the stomach through the lower esophageal sphincter did not present a problem because the pressure gradually increased in the esophagus despite any air leak. www.giejournal.org
After the endoscopic examination was performed, endoscopic variceal flattening (intravariceal pressure) was determined from the endoscopic recordings by 2 independent investigators blinded to the pressure. The method that the investigators used to determine the initiation of variceal flattening was to identify complete collapse on the endoscopic videotape and then to rewind the tape to the point of the initiation of variceal flattening (Fig. 4). Pressure was then determined from the simultaneous esophageal lumen manometry readings. These pressures were measured 3 times for each varix, and the mean SD of the pressures was calculated and recorded as the intravariceal pressure. A Pearson correlation coefficient was also used to determine the agreement between observers.
Variceal wall tension measurements During the endoscopic procedure, a 20-MHz high-resolution US probe was used to image the distal portion of the esophageal varices. The probe was passed through the biopsy channel of the endoscope after the manometry catheter was removed. We did not measure size or the shape of the varices endoscopically. We measured the variceal circumference sonographically and calculated the variceal radius when the esophagus was at rest (resting variceal circumference and variceal radius). Measurement of the sonographic cross-sectional areas and the circumference of esophageal varices at rest are now considered the criterion standard for ‘‘size’’ measurements of esophageal varices.7,8 To identify the varices on the US image with the varices on the endoscopic image, the picturein-picture feature on the endoscope was used. Measurements of the radius and wall thickness of 10 varices in 3 patients with portal hypertension were performed on Image Pro Plus software (version 6.0, Media Cybernetics, Bethesda, Md) after the images were digitized. Volume 70, No. 3 : 2009 GASTROINTESTINAL ENDOSCOPY 409
Esophageal variceal pressure and wall tension
Vegesna et al
Figure 4. Endoscopic view of varices in vivo during air insufflation showing pre (A) and post (B) variceal flattening. Note that the varices are bulging into the lumen of the esophagus before air insufflation (A) and that the varices are collapsed after air insufflation (B).
RESULTS Model varix results
Figure 3. Endoscopic view of the varix model showing variceal collapse during air insufflation.
Two investigators blinded to each other’s readings and to the intravariceal pressures measured the radius and wall thickness of the varices. Wall thickness and variceal radius were measured from the high-resolution US images with Image Pro Plus software (Fig. 5). Variceal wall tension was calculated with the Laplace equation (Wall tension Z Change in pressure Radius of varix/Wall thickness). Variceal flattening pressure (intravariceal pressure) was measured by the air insufflation method described above for calculation of wall tension. 410 GASTROINTESTINAL ENDOSCOPY Volume 70, No. 3 : 2009
The mean (SD) intraluminal esophageal pressures at the initiation of model variceal flattening for the model varix at 5, 10, 15, 20, and 25 mm Hg pressure were 5.69 (0.34), 11 (0.32), 15.72 (0.51), 21.55 (0.63), and 25.8 (0.14) mm Hg, respectively. The correlation coefficient between the actual model variceal pressure and the measured pressure in the model esophagus at model variceal flattening was r Z 0.98. The correlation coefficient between the 2 readers was r Z 0.99. The mean 2 standard errors for both readers can be seen in Figure 6.
In vivo variceal pressure measurement A total of 10 varices in 3 individual patients were evaluated. The individual varices in each patient flattened at different pressures, indicating that each varix has a different intravariceal pressure associated with it. The mean (SD) for the varices in each subject was 12.16 (2.4), 23.2 (1.3), and 6.5 (2.2) mm Hg for subjects 1, 2, and 3, respectively. The interobserver variability between readers was r Z 0.99.
Wall tension The intravariceal pressure correlated well with the intravariceal radius (r Z 0.85), but not with wall thickness (r Z 0.35). The mean (SD) for the wall tension in subjects www.giejournal.org
Vegesna et al
Figure 5. High-resolution US image of the varices with picture-in-picture endoscopic view. The picture also shows the outlines of the varix circumference and the wall thickness. Note that the inner circle (T1) outlines the inner circumference of the varix. The outer ark (T2) outlines the outer circumference of the varix. The wall thickness and variceal radius can be derived from these measurements.
Esophageal variceal pressure and wall tension
Figure 6. Correlation between the esophageal varix pressures and measured esophageal pressures at the initiation of variceal flattening in the varix model.
Given the poor outcome associated with variceal bleeding, the identification of varices at high risk and the prevention of first bleeding episodes are critical objectives. Screening endoscopy is generally recommended for patients with cirrhosis to determine whether large esophageal varices are present. Currently, a combination of clinical and endoscopic findings, including an advanced Child-Pugh class of cirrhosis, large varices on endoscopy, and the presence of red wale markings on endoscopy, correlate with the risk of a first bleeding episode in patients with cirrhosis.9 Our group has also demonstrated that the total cross-sectional surface area of the varices10 and the pressure generated during peristaltic contraction can be used to predict future variceal bleeding.6 At least 4 minimally invasive methods of measuring intravariceal pressure have been developed that exploit the concept of variceal flattening.2-6 However, all of these methods require additional equipment that is not readily available in most endoscopy units. Air insufflation to flatten varices is a standard procedure performed by almost every endoscopist. In fact, most grading systems of esophageal varices use variceal flattening to define grade 1 varices. This is a rough grading system, given that a high enough pressure will flatten any untreated esophageal varix and that the air pressure or volume on most endoscopes can be adjusted (low, medium, and high). In this study, we used the concept of variceal flattening and quantified the flattening pressure by measuring the esophageal
lumen pressure at the initiation of variceal flattening. We first demonstrated this concept in a model varix system and found the results to be accurate and reliable. We then demonstrated the feasibility of performing this variceal pressure measuring method in patients with virgin esophageal varices. Finally, we combined this variceal pressure measuring method with measurements on high-resolution endoluminal US to calculate the variceal wall tension with the Laplace equation. Force balance laws explain why, at the point of variceal flattening, the esophageal lumen pressure is equal to the intravariceal pressure. With this new method, the pressure within the varix at variceal flattening can be unambiguously inferred from the measured esophageal lumen pressure. The variceal flattening pressures of the model esophagus were found to strongly correlate with the flattening pressures within the model varices during air insufflation. It is clear from the literature that portal pressures do not reflect or correlate with esophageal variceal pressures.3 In fact, one of the reasons a variceal pressure measuring device is needed is because we cannot extrapolate hepatic portal venous gas (HPVG) to variceal pressures directly. We believe that the pressure head within the portal vein is damped by the high-resistance venous circulation (the palisade vessels) when the blood has to pass through these vessels to get from the stomach to the esophagus. This was demonstrated in 2 previous studies.6,11 Therefore, a pressure reading of 6 mm Hg in the esophageal varices is consistent with an HPVG of 12 or higher, which is required for varices to form because the pressure force generated in the portal vein is dissipated in palisade esophageal varices. We cannot validate variceal pressure measurements in vivo by measuring the HPVG because HPVG does not
www.giejournal.org
Volume 70, No. 3 : 2009 GASTROINTESTINAL ENDOSCOPY 411
1, 2, and 3 was 57.86 (9.22), 156.37 (46.79), and 8.29 (2.61) mm Hg, respectively.
CONCLUSIONS
Esophageal variceal pressure and wall tension
Vegesna et al
correlate with the variceal pressure. The only way of currently measuring variceal pressure directly is with a needle puncture method or another noninvasive method that has been previously validated. In this day and age of band ligation, it is no longer considered acceptable or ethical to perform variceal needle puncture to validate a noninvasive method of measuring intravariceal pressure. Luckily, there are other methods that use exactly the same principle (variceal flattening). The accuracy of these methods had been confirmed and validated in the past by use of variceal needle puncture.2 We can therefore extrapolate these findings to our new method and, with a great deal of confidence, state that our method is at least as valid as these other methods. The equalization pressure measured at the time of initial variceal flattening will not change, regardless of leaking air through the gastroesophageal junction or the upper esophageal sphincter, and is independent of the flexibility and distensibility of the esophagus. By newtonian physics and force balance equations, at the initiation of variceal flattening, the pressure inside the varix equals the pressure outside the varix. This has been demonstrated by use of other noninvasive methods for measuring esophageal variceal pressure.2 In fact, the entire concept of using air for measuring intravariceal pressure is based on previously validated studies that used similar methods (variceal flattening).2-6 Although there is often movement in the esophagus, which can make pressure measurement more difficult, the actual flattening pressure does not change on the basis of the flexibility, distensibility, or movement of the esophagus. Because the variceal flattening pressure was measured with the esophagus at rest (resting esophageal pressure), at 1 instant in time by increasing the pressure within the esophageal lumen to above the closing pressure within the perforating veins, we essentially have negated the effect of flow within the perforating vessels at the time of variceal flattening. These small thin-walled vessels close under increased esophageal pressure at a much lower closing pressure than did the large varices that were measured. In addition, by use of this new method it is easy to repeat the pressure measurements over and over again. We determined variceal wall tension in vivo by measuring variceal radius and wall thickness with high-resolution EUS on the basis of the previous work of Schiano et al7 and found a strong correlation between wall tension and variceal radius, but not with wall thickness. These same findings were shown previously by Jackson et al12 with the needle puncture method to measure variceal pressure and high-resolution EUS to measure the radius and wall thickness of the varices. The lack of correlation between wall thickness and variceal size in this study implies that variceal wall tension cannot be accurately estimated by the endoscopic observation of variceal size alone or even by size measurements combined with intravariceal pressure measurements.
We believe that the measurement of the resting variceal pressure has potential utility in evaluating the effectiveness of pharmaceutical therapy in individual patients and in evaluating potential medications that can lower variceal pressure in research studies. In fact, prior investigators have shown a decrease in the resting variceal pressure with b-blocker therapy, thus demonstrating the effectiveness of b-blockers.13 In summary, given the poor outcome associated with variceal bleeding, the identification of varices at high risk and the prevention of first bleeding episodes are critical objectives. Standard endoscopy with air insufflation and manometry can be used as an accurate, simple, and reproducible method to measure intravariceal pressure in a varix model system. This method was tested in a clinical setting to determine the feasibility of measuring intravariceal pressure. Finally, a noninvasive method of calculating variceal wall tension was developed and tested. This new technique may help provide a more accurate method for determining the risk of initial variceal bleeding and for monitoring pharmacologic therapy. We believe that the potentially wide applicability and ease of use make this method of high interest to clinicians and investigators.
412 GASTROINTESTINAL ENDOSCOPY Volume 70, No. 3 : 2009
www.giejournal.org
REFERENCES 1. Kleber G, Sauerbruch T, Fischer G, et al. Pressure of intraesophageal varices assessed by fine needle puncture: its relation to endoscopic signs and severity of liver disease in patients with cirrhosis. Gut 1999;30:228-32. 2. Bosch J, Bordas JM, Rigau J, et al. Non-invasive measurement of the pressure of esophageal varices using an endoscopic gauge: comparison with measurements by variceal puncture in patients undergoing endoscopic sclerotherapy. Hepatology 1986;6:667-72. 3. Rigau J, Bosch J, Bordas JM, et al. Endoscopic measurement of variceal pressure in cirrhosis: correlation with portal pressure and and variceal hemorrhage. Gastroenterology 1989;96:873-80. 4. Gertsch P, Wheatley AM, Maibach R, et al. Experimental evaluation of an endoscopic balloon for manometry of esophageal varices. Gastroenterology 1991;101:1692-700. 5. Miller LS, Dai Q, Thomas A, et al. A new ultrasound-guided esophageal variceal pressure measuring device. Am J Gastroenterol 2004;99:1267-73. 6. Miller LS, Kim JK, Dai Q, et al. Mechanics and hemodynamics of esophageal varices during peristaltic contraction. Am J Physiol Gastrointest Liver Physiol 2004;287:G830-5. 7. Schiano TD, Adrain AL, Cassidy MJ, et al. Use of high-resolution endoluminal sonography to measure the radius and wall thickness of esophageal varices. Gastrointest Endosc 1996;44:425-8. 8. 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;24:552-5. 9. North Italian Endoscopic Club for the Study and Treatment of Esophageal Varices. Prediction of first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices: a prospective multicenter study. N Engl J Med 1988;319:983-9. 10. Miller L, Banson FL, Bazir K, et al. Risk of esophageal variceal bleeding based on endoscopic ultrasound evaluation of the sum of esophageal variceal cross-sectional surface area. Am J Gastroenterol 2003;98:454-9. 11. Schiano DT, McCray HW, Liu J-B, et al. In vivo comparison of esophageal varices at and above the diaphragmatic high pressure zone using high resolution endoluminal sonography. J Clin Gastroenterol 1998;26:249-52.
Vegesna et al 12. Jackson WF, Adrain AL, Black M, et al. Calculation of esophageal variceal wall tension by direct sonographic and manometric measurements. Gastrointest Endosc 1999;50:247-51. 13. Nevens F, Sprengers D, Feu F, et al. Measurement of variceal pressure with an endoscopic pressure sensitive gauge: validation and effect of propranolol therapy in chronic conditions. J Hepatol 1996;24: 66-73.
www.giejournal.org
Esophageal variceal pressure and wall tension
Received June 30, 2008. Accepted November 12, 2008. Current affiliations: Temple University Hospital, Philadelphia, Pennsylvania, USA. Reprint requests: Larry S. Miller, MD, Section of Gastroenterology, 8th Floor, Parkinson Pavilion, Temple University Hospital, Philadelphia, PA 19140.
Volume 70, No. 3 : 2009 GASTROINTESTINAL ENDOSCOPY 413