Over the last two decades, our understanding of the treatment and prognosis of portal hypertension has continued to improve.1, 2 The use of techniques to measure the wedged hepatic venous pressure (WHVP), developed more than 50 years ago, has played a major role in elucidating the pathophysiology of the syndrome3 and, consequently, in developing currently available pharmacologic therapy. Recently, a role for measurements of WHVP has also been proposed in evaluating the progression of chronic liver diseases.4 This measurement has been shown to be the best predictor of the development of complications of portal hypertension in patients with early cirrhosis.5 Investigators in this area of research need not be convinced about the importance of the WHVP technique in advancing our knowledge of portal hypertension from the experimental arena3 to current clinical applications.6–8 However, there is an unanswered question that has lingered with us for more than 20 years: Should measurement of WHVP be incorporated into clinical practice?
To begin to answer this question, we must assure ourselves that the technique is executed properly. In a recent study performed to evaluate a new pharmaceutical agent, one of us (R.J.G.) was asked to be a blind reviewer of WHVP tracings performed by centers experienced in the measurement of WHVP. Even though minimal criteria for acceptable measurements have been established (Table 1), approximately 30% of studies% had to be rejected as the tracings obtained were uninterpretable. Our experience is not unique (Jaime Bosch, personal communication). We cannot draw conclusions about the usefulness of this technique in a clinical setting if investigational studies designed to answer the question do not themselves comply with minimal criteria for technical adequacy. To achieve results that are consistent and comparable from center to center, meticulous attention to detail is required (Table 1).
|1. A recorder capable of producing a permanent tracing of pressure values.|
|2. A quartz pressure transducer that can detect changes in venous pressure (not arterial!).|
|3. An occlusion balloon catheter.|
|Adequate calibration and recording|
|1. Use an appropriate scale. Venous pressures have an upper range of approximately 30–40 mm Hg. Therefore, scales used for arterial pressure measurements are not adequate. To be able to detect small changes, the scale should be set at 1 mm Hg = 1 mm on the scale (or more, e.g., 1 mm Hg = 2.5 mm on the scale; Fig. 2).|
|2. Use slow recording speed. Stabilization of venous pressures should be evaluated over a period of approximately 1 min for WHVP or 15 s for FHVP. The appropriate speed is <5 mm/s, optimally 1–2 mm/s. Note that in a “normal” ECG with a speed of 25 mm/s, one page of tracing includes approximately 10 s of measurement and this is not adequate for accurate interpretation of the tracing.|
|3. Check the accuracy of the transducer calibration by obtaining tracings of a known external pressure (e.g., a column of water of 13.6 cm should read 10 mm Hg and 27.2 cm H2O should read 20 mm Hg). If a transducer does not calibrate exactly against a known external pressure, replace it.|
|4. Place the transducer at the level of the right atrium (midaxillary line). The intravascular pressures will read higher if the transducer is lowered, but they will decrease if the transducer is raised. Record the IVC pressure on the tracing at the level of the liver (hepatic veins) before catheterizing the hepatic vein. Catheterize preferably the main right hepatic vein.|
|1. Do not advance the catheter too far into the hepatic vein when measuring the free pressure (or FHVP). The FHVP should not be more than 1 mm Hg greater than the IVC pressure. Greater differences require withdrawal of the catheter closer to the IVC for an accurate measurement of FHVP.|
|2. Record the tracing for 45–60 s to allow the measure to stabilize. Some patients may require a longer time! Also, continue recording when deflating the balloon to recheck the FHVP.|
|3. Obtain a mean pressure.|
|4. Repeat measurements at least three times to make sure that values obtained are reproducible. If they are not, check the wedged position of the catheter.|
|5. Check the inflated balloon for total occlusion of the hepatic vein (Fig. 1). If it is not (e.g., venous-to-venous shunts, insufficient inflation of the balloon), the measurements should be repeated either by moving the balloon catheter distal to the venous-to-venous shunts, or, when the drainage is to another hepatic vein, by changing the position of the catheter to another hepatic vein without venous-to-venous shunts. A hepatic vein that drains into another hepatic vein or distal to the balloon occlusion will underestimate the WHVP. In rare cases, the measurement cannot be accomplished. Checking for total occlusion of the balloon (wedged position) should be performed at the end of the measurement by slowly injecting 5 mL of contrast into the hepatic vein while the balloon is inflated. This should show the typical wedged (sinusoidal) pattern and no communication with other hepatic veins. After deflating the balloon the dye should wash out quickly. Do this consistently (i.e., do not check for balloon occlusion before measurements on one procedure and after the measurements in another).|
|6. If the patient is premedicated, for comparative purposes, subsequent measurements should be performed under the same conditions.|
|7. Register on tracing ongoing events. For example, cough or slight movements cause artifacts that may give inaccurate readings. Any such changes that affect the measurement should be noted on the tracing.|
|8. Never rely on digital readings on the screen! These are instantaneous readings and may not be representative (because of cough, movement) of the correct measurement. Digital readings should not be accepted owing to the lack of a hard copy that can be reviewed.|
Direct measurement of portal venous pressure (PVP) is invasive and inconvenient. In 1951, Myers and Taylor9 first described WHVP, which is the measurement of the sinusoidal pressure, an indirect measurement of PVP. Since then, WHVP has been shown to be very safe and the rate of successful hepatic vein catheterization is greater than 95%. By threading a small catheter into a hepatic vein until it cannot be advanced further, a “wedged” hepatic venous pressure is obtained. As the hepatic vein is occluded, a continuous column of fluid between the catheter and the sinusoid is formed, resulting in a pressure reading that is equal to the sinusoidal pressure (Fig. 1). In normal livers, the low-resistant sinusoidal network dissipates most of the pressure back up from the wedged catheter. As there is no direct connection, via a static column of fluid, between the catheter and the portal tributaries, the pressure reading of the transducer reflects sinusoidal pressure (in normal livers is slightly lower than portal pressure). This is also observed in presinusoidal causes of portal hypertension such as schistosomiasis or early primary biliary cirrhosis. Because the catheter in these cases is not in continuity with the area of increased resistance, the recorded pressure will be that of the normal sinusoids and not of the increased pressure in the portal vein. In these cases, WHVP will be an underestimation of the PVP.11, 12 Conversely, in alcoholic cirrhosis and in most cases of hepatitis C and B virus-induced cirrhosis, connections between sinusoids are decreased because of narrowing of the sinusoidal vascular bed (caused by collagen deposition in the space of Disse, compression by regenerative nodules, and microthrombosis). As there is little dissipation of pressure in the narrowed sinusoids, the static column of blood extends from the catheter to the portal vein and the WHVP is virtually equal to the PVP (Fig. 1).3, 11–14
Currently, the most commonly utilized parameter is not the WHVP, but hepatic venous pressure gradient (HVPG), the difference between WHVP and free hepatic venous pressure (FHVP). HVPG represents the gradient between the portal vein and the intraabdominal vena caval pressure. Whereas both the WHVP and FHVP are affected equally by intraabdominal pressure, their gradient, HVPG, is not. Unlike PVP or WHVP, which can be elevated falsely in the presence of ascites and elevated intraabdominal pressure, the measurement of HVPG incorporates its own zero reference point and is not affected by increases in intraabdominal pressure. Furthermore, the use of the HVPG eliminates another very important source of error, the external zero reference point, which may vary from center to center.
A recent study indicates differences in the values obtained when the catheter is wedged in different hepatic veins, which is a cause of concern.15 These differences probably may be due to the heterogeneity of sinusoidal involvement by the disease process affecting the liver. The smaller the vein where the catheter is wedged, the greater the potential discrepancy with a subsequent measurement in a different hepatic vein. This observation reflects the finding that cirrhosis is not a homogeneous disease.16, 17
The use of a balloon-tipped catheter18 (Fig. 1) versus a straight catheter provides both theoretical and practical advantages. The balloon catheter occludes a larger hepatic venous branch and can, theoretically, measure WHVP over a wider vascular territory of the liver compared with the straight catheter. The use of balloon catheter also provides a practical advantage as it allows repeated measurements from the same hepatic vein and avoids the decompressive effect of venous-to-venous shunts that are proximal to the balloon. Conversely, a straight catheter has to be advanced and withdrawn for each WHVP and FHVP measurement, making it difficult to wedge the same venule with each successive pressure determination.
Having experience with this procedure for more than 25 years, we are confident that properly executed, the balloon catheter technique produces HVPG values that are accurate and reproducible. Adherence to guidelines suggested in Table 1 is extremely important. Furthermore, pressure tracings should be annotated and recorded so that they can be reviewed subsequently by an independent observer (Fig. 2).
In summary, we do not believe that proponents of HVPG measurement have overemphasized its importance. Even if this method does not achieve the clinical utility that many of us believe it should, we have no doubt of the importance that it played and is still playing in advancing our knowledge of the portal hypertensive syndrome. What is also clear to us is that to reach any conclusion about this technique, one must first ensure that the technique is performed properly.