| Background and aimsVariceal pressure has been identified as the key factor leading to variceal rupture. Measuring variceal pressure is important in predicting esophageal variceal bleeding. Noninvasive balloon techniques assume that varices behave as an elastic structure because of their thin walls and lack of external tissue support; thus, the pressure needed to compress a varix (which can be sensed under direct vision using clear balloons) equals the pressure inside the varix. Manometry uses an endoscopic balloon to measure variceal pressure. Up until now, this method relied on the visual appearance of the varices and, therefore, was subjected to observer bias. Endoscopic balloon methods of measuring variceal pressure have not gained wide popularity. We have recently developed a computerized endoscopic balloon manometry (CEBM), with computer visualization to determine the moment of variceal wall collapse during balloon manometry instead of the visual evaluation of variceal compression used in the traditional balloon method. In this paper, we report the preliminary results of in-vitro and in-vivo testing of CEBM.The aims of the present study were: (1) to develop a new computerized endoscopic balloon manometry; (2) to validate this technique of CEBM in-vitro; (3) to test the clinical reliability and feasibility of this manometry by compared with the HVPG; and to correlate measured pressure values to endoscopic bleeding risk parameters as previously described; (4) to assess the applicability of this method for predict a first variceal hemorrhage in cirrhotic patients during the prospective study, and to detect clinical factors in the evaluation of the risk of bleeding. Patients and methodsThe CEBM system, comprising an esophageal variceal manometer and a computer, records variceal pressure and manometry images simultaneously. (1) In the in-vitro study, variceal models were fixed inside an artificial esophagus, into which an endoscope with transparent balloon was inserted for intraluminal pressure measurement. The artificial varix was filled with water and connected to a water column to modulate the intraluminal pressure. This CEBM system was tested blindly in three variceal models (the artificial variceal models, which were 3, 6, and 8mm in diameter) with different intraluminal pressures, ranging from 8 to 36 cmH2O. (2) CEBM was also used to measure variceal pressure in 23 patients with liver cirrhosis and esophageal varices, and the results were compared with the hepatic venous pressure gradient (HVPG), to assess the accuracy of this method. (3) Fifty-seven patients with liver cirrhosis and esophageal varices who had never experienced variceal bleeding were followed for 12 months. The patients underwent variceal pressure measurement by CEBM technique. The endpoint of the study was the presence or absence of a variceal hemorrhage. The relation between variceal hemorrhage with age, sex, etiology of cirrhosis, endoscopic findings of varices (size of varices, red color signs), variceal pressure, Child-Pugh's class, and ascites were studied.Results(1) The experiments proved that the measured pressure by CEBM has high stability. Variceal pressure measurements were technically satisfactory in-vitro and in clinical study. (2) In-vitro study, the measured intraluminal pressure correlated significantly with the actual intraluminal pressure for different diameters (r≥0.993, P <0.001). No obvious measurement bias was found for any of the varices (95%C.I were -0.13 cmH2O to 0.33 cmH2O). The diameter of artificial varices did not affect the measurement of variceal pressure. (3) Variceal pressure measurements with CEBM were technically successful in 23 patients. The endoscope with the deflated balloon could be introduced into the stomach without difficulty. No adverse effect related to the measuring procedure and sedation was observed, except for slight retching during intubation. We noticed that in particular there was no retrosternal pain, variceal bleeding, or uncontrolled coughing during slow inflation of the balloon. The time required to make five measurements of variceal pressure in each patient was 12±3 minutes (mean±SD). There was a close agreement between the two individuals regarding the point of flattening of the balloon variceal markers (correlation coefficients: r≥0.998). Overall, the value for mean variceal pressure (25.6±4.8 mmHg) was higher than that for mean HVPG (18.1±3.7 mmHg, P<0.001). Variceal pressure values (28.9±2.9 mmHg) were higher in patients with bleeding compared with nonbleeding patients (21.3±3.0 mmHg, P<0.001). Regression analysis showed a good correlation between variceal pressure measured with CEBM and the HVPG (r= 0.858, P <0.001). However, determination of the variceal pressure was not very satisfactory in two patients. The presence of previous bleeding episodes was strongly associated with higher variceal pressure (28.6±3.1 mmHg vs 21.3±3.0 mmHg, P<0.001). Variceal pressure correlated positively to variceal sizes (large varix, 26.6±4.7 mmHg vs small varix, 22.6±3.8 mmHg, P=0.05). We also found a significant relationship between variceal pressure and the presence of red color signs (27.3±4.2 mmHg vs 22.5±4.1 mmHg, P=0.015). Child–Pugh class parameter did not correlate with variceal pressure. The average of variceal pressure in Child–Pugh class A (n=11) was 25.09±3.94 mmHg, the average of variceal pressure in Child–Pugh class B plus C (n=12) were 25.71±5.52 mmHg (t=0.31, P=0.76). (4) In all 57 cirrhotic patients, thirty-four patients (34/57, 59.6 %) developed a variceal hemorrhage. In univariate analysis, the level of variceal pressure (bleeder 28.87±2.61 mmHg; nonbleeder 20.43±2.98 mmHg; P < 0.001), the diameter of varices (bleeder 8.91±2.04 mm; nonbleeder 7.09±2.75 mm; P = 0.006), and the endoscopic red color sign on the variceal wall (bleeding rate, bleeder 81% vs nonbleeder 47.2%, P = 0.012) predicted a higher risk of variceal hemorrhage. With the results obtained from univariate analysis, a multiple logistic regression model was created which revealed that variceal pressure was major predictor of the risk for a first variceal bleeding (OR = 2.817, P = 0.003, 95.0% C.I. were 1.437 to 5.521). The area under the receiver operating characteristic (ROC) of variceal pressure for predicting variceal bleeding was 0.98 and at variceal pressure cutoff value of 25.3 mmHg, specificity and sensitivity were 91%. Higher variceal pressures have been documented in patients with large varices(gradeⅢ26.56±4.46 mmHg vs gradeⅡ21.74±4.49 mmHg, P=0.02), and in those with red color signs ( bleeder, 28.40±2.41 mmHg vs 23.25±3.69 mmHg, P < 0.001; nonbleeder, 26.74±4.95 mmHg vs 19.97±2.18 mmHg, P < 0.001), which are those more prone to bleed.ConclusionsOur preliminary results indicate that CEBM of esophageal varices is feasible and accurate. This technique may become a more reliable method for noninvasive measurement of variceal pressure and warrants further investigation. The level of variceal pressure is a major predictor for variceal bleeding in cirrhotic patients.
|
PDF Full Text Request