SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

Background : The acute effects of β-blockers may be different from chronic; mechanisms underlying this difference are poorly elucidated.

Aim : To assess portal pressure and its pathophysiological determinants after acute and chronic administration of nadolol.

Methods : In 24 patients with cirrhosis and portal hypertension hepatic venous pressure gradient, portal blood flow and resistance to portal blood flow were measured before, 60–90 min after acute administration of nadolol, and after 1 month. Patients were good-responders if hepatic venous pressure gradient was ≤12 mmHg, or decreased by at least 20%.

Results : Eleven and 13 patients were good- and poor-responders to acute administration, respectively. Acute poor-responders showed a lower decrease in portal blood flow (P = 0.04) and a less evident decrease in mean arterial pressure (P < 0.001). Eleven and 13 patients were good- and poor-responders to chronic administration, respectively. Chronic poor-responders showed a larger increase in resistance to portal blood flow compared with good-responders (P = 0.01).

Disagreement between acute and chronic effects was seen in 12 patients: six were acute good-responders chronic poor-responders and six were acute poor-responders chronic good-responders. Acute good-responders chronic poor-responders patients had the smallest decreases in portal blood flow and in mean arterial pressure after acute administration, while acute poor-responders chronic good-responders showed the largest (P = 0.05 and 0.01).

Conclusions : Disagreement between acute and chronic effects of nadolol on hepatic venous pressure gradient is common. The mechanism responsible is complex, the acute effect being mainly modulated by arterial hypotension and the chronic effect by changes in portal resistance.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

It was already observed that portal hypotensive effect of β-blockers after a single acute administration may be markedly different from that obtained during chronic administration.1 These differences may have important practical implications, because chronic effect of β-blockers is predictive of clinical effectiveness of treatment,2–6 while acute effect is not.7 Recently, it was reported a fair agreement between acute effects of propranolol and chronic effects of nadolol (N) in a series of patients undergoing haemodynamic evaluation.8, 9 Therefore, the interest in the relationships between acute and chonic effects has been fostered.

In the present study, we aimed at investigating if there are differences among acute and chronic effects of the non-selective β-blocker, N, on hepatic venous pressure gradient (HVPG), and to investigate possible mechanisms for this difference. Indeed, on a a priori basis, it may be hypothesized that differences between the acute and chronic administrations may arise from a different time course of the haemodynamic changes, or from a different response of the main determinants of portal pressure [namely portal blood flow (PBF) and resistance to portal blood flow (RPBF)] to drug administration. Furthermore, different changes in the determinants of portal pressure may be due to different systemic haemodynamic changes occurring after acute or chronic administration of β-blockers.

To investigate these points, we evaluated HVPG, PBF, RPBF and systemic haemodynamics in patients with cirrhosis and portal hypertension before, after the first administration of N, and after 1 month of chronic treatment.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

Twenty-four patients with cirrhosis and portal hypertension, who were referred to our liver haemodynamic laboratory for evaluation of portal hypertension, took part in this study. Twelve patients had alcoholic cirrhosis, 11 hepatitis C virus (HCV)-related and one hepatitis B virus (HBV)-related cirrhosis. In all cases diagnosis was based on liver histology; in all histological examination showed micronodular cirrhosis. There were 15 men and nine women; mean age was 56 years. No patient had previous variceal bleeding, or had had previous treatments for portal hypertension. Five patients with ascites were treated with potassium kanrenoate for the whole study period. Five patients were on lactulose for previous episodes of hepatic encephalopathy. Main clinical and laboratory data at inclusion in the study are given in Table 1. The study protocol conformed to the Declaration of Helsinki, and was approved by the competent Ethics Authorities. Consent to the study was obtained in writing by all participating subjects.

Table 1.  Main clinical, biochemical and endoscopic data in the 24 patients with cirrhosis
  1. HCV, hepatitis C virus; HBV, hepatitis B virus.

Sex (M/F)15/9
Age (years)56 ± 5
Aetiology (alcoholic/HCV/HBV)12/11/1
Ascites5
Encephalopathy1
Serum creatinine (μm)61 ± 8
Child-Pugh class (A/B/C)12/10/2
Variceal size (F1/F2/F3)0/13/11
Red weal marks (−/+)7/17

Procedure

After overnight fast, the patients were brought to the haemodynamic laboratory, and duplex Doppler examination of the portal vein was first performed according to a previously described methodology10 using a Toshiba Sonolayer SSA-270 apparatus (Toshiba, Tokyo, Japan) with colour Doppler sonography and a 3.75 MHz sector electronic probe. Portal blood flow was calculated according to the following formula:11

  • image

In our laboratory, coefficient of variation of PBF for replications after 1 month was 11.9 ± 9.2%, as determined in nine control subjects.12

Immediately afterwards, hepatic vein catheterization was performed using a balloon catheter, according to a procedure which was previously described.13 In brief, a sheath introducer was placed in the right femoral vein under local anaesthesia, and a balloon catheter (Meditech F7, Watertown, MA, USA) was advanced under fluoroscopic control till a main hepatic vein (usually the right hepatic vein). HVPG, index of the portal pressure in conditions of sinusoidal or postsinusoidal portal hypertension,14, 15 was calculated as occluded minus free hepatic venous pressure. In all cases three measurements were performed, electronic means were recorded, and results were expressed as the mean of the three values. Permanent tracings of the pressure measurements were always obtained. Resistance to PBF was calculated as HVPG divided by PBF and expressed as mmHg × min/L. During the whole procedure, electrocardiogram (ECG), heart rate (HR) and arterial blood pressure were monitored.

After completion of all baseline measurements, all patients received a single oral dose of N 80 mg, and haemodynamic measurements were done again after 60–90 min. In a preliminary investigation in five patients observed for up to 180 min after acute oral administration of N 80 mg, HR, mean arterial pressure (MAP) and HVPG were stable from 60 to 180 min; therefore the time frame 60–90 was chosen to replicate haemodynamic measurements, in order to limit the duration of the invasive procedure.

Patients then received a chronic N treatment at a dose reducing resting HR approximately by 25% (mean dose 72 mg/day), and were followed in the out-patient clinic. After 1 month of effective β-blockade, they were admitted again and studied by hepatic vein catheterization and duplex Doppler ultrasonography. Procedures were performed in the morning approximately at the same hour with the same methodology, in fasting patients who were given the usual morning dose of N.

During the whole evaluation period, patients were seen fortnightly as out-patients. Compliance was assessed according to anamnesis and HR measurements. Blood levels of N were not measured. No patient reported side-effects requesting withdrawal of the drug.

Definition of agreement/disagreement between acute and chronic effects

Because decrease in HVPG by 20% or more, or decrease in HVPG to 12 mmHg or less are considered predictors of good response to treatment,2, 4–6 results obtained in patients reaching (or not reaching) these haemodynamic end points after acute and chronic administration were considered in agreement. The others were considered in disagreement. They were divided into two groups, i.e. ‘acute good-responders but chronic poor-responders’ (AGCP), when patients reached criteria for good response after acute administration, but this was not kept during chronic administration and ‘acute poor-responders but chronic good-responders’ (APCG) when criteria for good response were only obtained after chronic administration.

Statistics

Results are given as mean ± S.E.M. Comparisons were made by Wilcoxon's test or Mann–Whitney's test, when applicable. Multiple comparisons between groups were tested by Kruskal–Wallis anova and z-statistics correction. Correlations were investigated by the Spearman rank correlation coefficient. The null hypothesis was rejected at the 0.05 probability level.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

Whole series haemodynamic changes

In the whole series baseline HVPG was 18.9 ± 0.5 mmHg, and decreased significantly to 15.8 ± 0.7 after acute administration (P < 0.001) and to 16.2 ± 0.5 after chronic administration (P < 0.001). Mean decrease was 17% and 14%, respectively. Baseline PBF was 1009 ± 68 mL/min, and decreased to 910 ± 72 after acute administration (P = 0.01) and to 740 ± 47 after chronic administration (P < 0.001). Mean decrease was 10% and 24%, respectively. Baseline RPBF was 20.9 ± 1.7 mmHg × min/L, and slightly decreased after acute administration (19.9 ± 1.80; P < 0.05; mean decrease = 5%), but increased significantly after chronic administration (24.3 ± 1.9; P = 0.01; mean increase = 20%).

Baseline HR was 83 ± 2/min, and decreased to 64 ± 2 after acute administration (P < 0.001; mean decrease = 23%) and to 60 ± 1 after chronic administration (P < 0.001; mean decrease = 28%). Baseline MAP was 105 ± 2 mmHg, and decrease to 95 ± 2 after acute administration (P < 0.001; mean decrease = 9%), and to 98 ± 2 after chronic administration (P < 0.01; mean decrease = 6%).

Haemodynamic changes in acute responders/non-responders

After acute administration of N, 11 patients met the criteria for a good response and 13 failed. The two groups were not different for demographic, clinical, biochemical, or endoscopic variables. Acute good-responders showed a significant decrease in PBF (from 943 ± 88 to 789 ± 97 mL/min; P < 0.05) and in RPBF (from 21.3 ± 2.3 to 19.5 ± 2.3 mmHg × min/L; P < 0.05), while acute poor-responders showed minor and non-significant changes in both determinants of portal pressure, and percentage decrease in PBF was higher in good- than in poor-responders (−17 ± 4 vs. −4 ± 5%; P < 0.05). In addition, acute good-responders showed a more pronounced percentage decrease in MAP (−15 ± 2 vs. −4 ± 1%; P = 0.001), while changes in HR were nearly identical. In the whole series percentage change in HVPG was significantly correlated to changes in PBF (rs = 0.47; P < 0.05) and to changes in MAP (rs = 0.63; P < 0.01).

Haemodynamic changes in chronic responders/non-responders

After chronic administration of N, 11 patients met the criteria for a good response and 13 failed. The two groups were not different for demographic, clinical, biochemical, or endoscopic variables. Both groups showed significant decrease in PBF (P < 0.01) of the same size, but chronic poor-responders showed a significant increase in RPBF (from 19.6 ± 1.7 to 25.6 ± 2.7 mmHg × min/L; P < 0.01) averaging a 32% increase, while chronic good-responders only showed minor and non-significant changes. HR and MAP changes were nearly identical in both groups. No significant correlation was found between changes in HVPG and changes in PBF or MAP.

Agreement/disagreement between acute and chronic changes

Of the 11 acute good-responders, five were also chronic good-responders (AGCG), and six were chronic poor-responders (AGCP). Of the 13 acute poor-responders, seven were also chronic poor-responders (APCP), and six were APCG. Baseline demographic, clinical, biochemical, endoscopic and haemodynamic variables were not significantly different among the four groups of patients. When the changes induced by the treatment were compared among the four groups, it resulted (Table 2) that an acute decrease in PBF was more pronounced in AGCP than in APCG (P = 0.05), and a lack of increase in RPBF after chronic administration was observed only in APCG (P = 0.02). AGCP patients also showed a significantly larger increase in RPBF after chronic than acute treatment (P < 0.05). As for the systemic haemodynamic changes, APCG responders were characterized by a lesser decrease in HR after acute administration (P = 0.02), and AGCP by a more marked acute decrease in MAP (P = 0.02), and a significantly lesser percentage decrease in MAP after chronic than acute treatment (P < 0.05). Individual results in AGCP and APCG patients are reported in Figures 1 and 2.

Table 2.  Haemodynamic variables in the four groups of patients classified according to the HVPG response (P-value according to Kruskal–Wallis anova and z-statistics for multiple comparisons)
VariableAPCPAPCGAGCPAGCGP-value
  1. * APCG different from APCP or AGCP (P = 0.05).

  2. † APCG different from APCP (P = 0.05).

  3. ‡ AGCP different from APCP or APCG (P = 0.02).

  4. § APCG different from AGCP (P = 0.07).

  5. ¶ Different from acute change (P < 0.05, Wilcoxon test).

  6. HVPG, hepatic venous pressure gradient; PBF, portal blood flow; RPBF, resistance to portal blood flow; HR, heart rate; MAP, mean arterial pressure; APCP, acute poor-responders but chronic poor-responders; APCG, acute poor-responders but chronic good-responders; AGCP, acute good-responders but chronic poor-responders; AGCG, acute good-responders but chronic good-responders.

Baseline HVPG18.1 ± 1.020.9 ± 1.118.0 ± 1.018.5 ± 1.00.18
 HVPG after acute administration17.0 ± 1.118.4 ± 0.813.5 ± 1.013.6 ± 1.30.01
 HVPG after chronic administration17.3 ± 1.115.8 ± 0.516.8 ± 0.814.5 ± 0.90.28
 Percentage change in HVPG after acute administration−6 ± 2−11 ± 1¶−25 ± 2¶−27 ± 40.001
 Percentage change in HVPG after chronic administration−5 ± 1−24 ± 2¶−7 ± 3¶−22 ± 10.001
Baseline PBF1010 ± 1031129 ± 193968 ± 123912 ± 1400.77
 PBF after acute administration921 ± 891121 ± 190740 ± 90848 ± 1950.29
 PBF after chronic administration717 ± 62895 ± 122718 ± 75612 ± 1040.30
 Percentage change in PBF after acute administration−8 ± 6¶+1 ± 7§−24 ± 2§−11 ± 70.05
 Percentage change in PBF after chronic administration−26 ± 7¶−17 ± 5−24 ± 5−31 ± 70.53
Baseline RPBF19.1 ± 2.522.3 ± 4.720.2 ± 2.622.7 ± 4.30.95
 RPBF after acute administration20 ± 4.419.7 ± 4.019.5 ± 2.319.5 ± 4.50.98
 RPBF after chonic administration26.1 ± 4.419.9 ± 3.224.9 ± 3.426.3 ± 4.70.69
 Percentage change in RPBF after acute administration+5 ± 8¶−10 ± 5−2 ± 4¶−17 ± 50.08
 Percentage change in RPBF after chronic administration+40 ± 18*¶−7 ± 5*+24 ± 5*¶+20 ± 160.02
Baseline HR81 ± 383 ± 284 ± 486 ± 50.72
 HR after acute administration59 ± 2.468 ± 265 ± 465 ± 30.18
 HR after chronic administration58 ± 2.160 ± 362 ± 261 ± 40.52
 Percentage change in HR after acute administration−27 ± 2†−18 ± 2†−22 ± 2−25 ± 20.02
 Percentage change in HR after chronic administration−28 ± 4−28 ± 2−26 ± 1−29 ± 30.89
Baseline MAP100 ± 2102 ± 2110 ± 6108 ± 50.28
 MAP after acute administration95 ± 3100 ± 289 ± 596 ± 40.20
 MAP after chronic administration97 ± 494 ± 3103 ± 698 ± 50.58
 Percentage change in MAP after acute administration−5 ± 2‡−2 ± 1‡−19 ± 1‡−11 ± 20.001
 Percentage change in MAP after chronic administration−3 ± 4−8 ± 3−7 ± 2¶−9 ± 50.59
image

Figure 1. Individual values of hepatic venous pressure gradient (HVPG), mean arterial pressure (MAP), portal blood flow (PBF) and resistance to portal blood flow (PBF) in acute poor-chronic good-responders (APCG) to nadolol. Boxes represent mean values and S.E.M.

Download figure to PowerPoint

image

Figure 2. Individual values of hepatic venous pressure gradient (HVPG), mean arterial pressure (MAP), portal blood flow (PBF) and resistance to portal blood flow (PBF) in acute good-chronic poor-responders (AGCP) to nadolol. Boxes represent mean values and S.E.M.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

In the overall series we observed that both after acute and chronic administration of N a similar decrease in HVPG occurred, but after acute administration the decrease was mainly linked to a decrease in PBF, without marked changes in RPBF, while after chronic administration the overall effect is the resultance of a further decrease in PBF, associated with a moderate increase in RPBF. Systemic haemodynamic changes was also different in part, because decrease in MAP was more pronounced after acute administration.

Both after acute and chronic administration, the haemodynamic response was not uniform among patients, because approximately half of them respond with an effective decrease in HVPG (good-responders). Good-responders were not different from poor-responders according to any baseline clinical, biochemical, or haemodynamic feature; therefore there was no way of predicting in the single case who will respond to N administration with an effective decrease in HVPG. No study examined so far the mechanisms involved in the different response to acute and chronic administration of β-blockers, because of the complexity of assessing at the same time HVPG, PBF and RPBF. Simultaneous assessment by hepatic vein catheterization and duplex Doppler ultrasonography is the less invasive method to evaluate these portal haemodynamic variables. Despite some criticisms on the reliability of duplex Doppler measurements,16 there is agreement that these measurements are adequately reproducible provided a strict protocol of data recording is followed.10

Differences between good- and poor-responders after acute administration

These differences appear to be linked to the different effect of the first dose of N on the pathophysiological mechanisms leading to portal hypertension, i.e. PBF and RPBF. Indeed, good-responders showed a larger decrease in PBF compared with poor-responders (−17% vs. 4%; P = 0.04), while RPBF did not show a different course in the two groups. In addition, good-responders showed a larger decrease in MAP after a first dose (−15% vs. −4%; P = 0.001), while changes in HR were equal in both groups. Finally, a significant correlation was found between changes in HVPG and changes in PBF and in MAP, respectively. These data suggest that the different course of HVPG after acute administration is related to a variable effect on PBF, which in turn is related to a variable arterial hypotensive effect occurring after the first dose of N.

Difference between good- and poor-responders after chronic administration

These differences also appear to be linked to the different effect of the drug on the pathophysiological mechanisms leading to portal hypertension, i.e. PBF and RPBF. Indeed, poor-responders after chronic administration was characterized by a much larger increase in RPBF compared with good-responders (+32% vs. +6%; P = 0.01), associated with a similar decrease in PBF. No difference in changes in MAP or HR was documented between good- and poor-responders. These results, which are in agreement with those reported in a previous study from our laboratory,17 confirm that the difference between good- and poor-responders to chronic administration lies in a different response of RPBF to chronic β-blocker administration.

Agreement/disagreement between acute and chronic response

Because differences between good- and poor-responders after acute or chronic administration of N seem to be prevalently related to different mechanisms and sites of action (β-1 at the cardiac site for acute response, and β-2 at portal and/or collateral sites for chronic response), it is not surprising that disagreement between a condition of good- or poor-responder was frequently observed between a first administration and a chronic one. Indeed, in only half of the patients (12 of 24) there was agreement between acute and chronic effect, while in the other half there was disagreement; in addition this disagreement was equally divided between patients who showed a good acute response but were unable to keep it in the chronic administration (AGCP – six patients), and patients who were poor-responders after acute administration but gained a good response during chronic administration (APCG – six patients). The analysis of the pathophysiological variables in the four groups of patients divided according to acute and chronic response (Table 2) confirmed the interpretation given above. Indeed, the smallest reduction of PBF after acute administration (1%) was found in APCG patients, while the largest one was found in AGCP patients (−24%; P = 0.05). Similarly, the smallest decrease in MAP after acute administration (−2%) was found in APCG patients, and the largest one was found in AGCP (−19%; P = 0.001). In addition, the two groups had a opposite course as far as chronic changes in RPBF, which were nearly unchanged in APCG (−7%) and markedly increased in AGCP (+24%; P = 0.05).

From a practical point of view, these data confirm that it is not possible to use acute changes in HVPG as predictors of chronic effect of N, and if a useful predictor of clinical response to N treatment is requested, it is necessary to assess the chronic effect. This observation is important, because it is generally suggested by practice guidelines that treatment with β-blockers is to be monitored according to the haemodynamic response.18–21

In conclusion, a good response to acute or chronic administration of N was seen in less than half of patients with cirrhosis and portal hypertension, but disagreement between acute and chronic effect of N was frequently observed. The mechanisms responsible for these differences is complex, because the acute effect is mainly modulated by the inhibition of the β-1 cardiac receptor, which is variable in the various subjects, and the chronic effect is mainly modulated by the inhibition of the β-2 receptor located in portal and collateral circulation, which is also variable in the various subjects.

Acknowledgement

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

This study was supported in part by the Italian Ministry of Education, University and Research, Rome, Italy (PRIN 2003; National Project ‘Hemodynamic alterations in cirrhosis’).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References
  • 1
    Vorobioff J, Picabea E, Villavicencio R, et al. Acute and chronic hemodynamic effects of propranolol in unselected cirrhotic patients. Hepatology 1987; 7: 64853.
  • 2
    Groszmann RJ, Bosch J, Grace N, et al. Hemodynamic events in a prospective randomized trial of propranolol vs. placebo in the prevention of the first variceal hemorrhage. Hepatology 1990; 99: 14017.
  • 3
    Sacerdoti D, Merkel C, Gatta A. Importance of the 1-month-effect of nadolol on portal pressure in predicting failure of prevention of rebleeding in cirrhosis. J Hepatol 1991; 12: 1245.
  • 4
    Feu F, Garcia-Pagan JC, Bosch J, et al. Relation between portal pressure response to pharmacotherapy and risk of recurrent variceal haemorrhage in patients with cirrhosis. Lancet 1995; 346: 10569.
  • 5
    Villanueva C, Balanzò J, Novella MT, et al. Nadolol plus isosorbide mononitrate compared with sclerotherapy for the prevention of variceal rebleeding. N Engl J Med 1996; 334: 16249.
  • 6
    Merkel C, Bolognesi M, Sacerdoti D, et al. The hemodynamic response to medical treatment of portal hypertension as a predictor of clinical effectiveness in the primary prophylaxis of variceal bleeding in cirrhosis. Hepatology 2000; 32: 9304.
  • 7
    Valla D, Jiron D, Poynard T, Braillon A, Lebrec D. Failure of hemodynamic measurements to predict recurrent gastrointestinal bleeding in cirrhotic patients receiving propranolol. J Hepatol 1987; 5: 1448.
  • 8
    Aracil C, Lòpez-Balaguer JM, Monfort D, et al. Hemodynamic response to beta-blockers and prediction of clinical efficacy in the primary prophylaxis of variceal bleeding in patients with cirrhosis. Hepatology 2003; 38 (Suppl. 1): 296A.
  • 9
    de Madaria E, Palazon JM, Sanchez J, et al. Monitoring of HVPG during pharmacological treatment: a comparison between two methods. J Hepatol 2005; 42 (Suppl. 2): 79.
  • 10
    Sabbà C, Merkel C, Zoli M, et al. Interobserver and interequipment variability of echo-Doppler examination of the portal vein: effect of a co-operative training program. Hepatology 1995; 21: 42833.
  • 11
    Moriyasu F, Ban N, Nishida O, et al. Clinical application of a ultrasonic duplex system in the quantitative measurement of portal blood flow. J Clin Ultrasound 1986; 14: 57988.
  • 12
    Bolognesi M, Sacerdoti D, Merkel C, Bombonato G, Enzo E, Gatta A. Effects of chronic therapy with nadolol on portal hemodynamics and on splanchnic impedance indices using Doppler sonography: comparison between acute and chronic effects. J Hepatol 1997; 26: 30511.
  • 13
    Merkel C, Bolognesi M, Bellon S, et al. Prognostic usefulness of hepatic vein catheterization in patients with cirrhosis and esophageal varices. Gastroenterology 1992; 102: 9739.
  • 14
    Lebrec D. Methods to evaluate portal hypertension. Gastroenterol Clin North Am 1992; 21: 4159.
  • 15
    Lin HC, Tsai YT, Lee FY, et al. Comparison between portal vein pressure and wedged hepatic vein pressure in hepatitis B-related cirrhosis. J Hepatol 1989; 9: 32633.
  • 16
    DeVries PJ, Van Hattum J, Hoekstra JBL, De Hooge P. Duplex Doppler measurements of portal venous flow in normal subjects: inter- and intra-observer variability. J Hepatol 1991; 13: 35863.
  • 17
    Merkel C, Sacerdoti D, Bolognesi M, et al. Hemodynamic evaluation of the addition of isosorbide-5-mononitrate to nadolol in cirrhotic patients with insufficient response to the beta-blocker alone. Hepatology 1997; 26: 349.
  • 18
    Groszmann RJ, Bendtsen F, Bosch J, et al. Baveno II consensus statements: drug therapy for portal hypertension. In: De Franchis, R, ed. Portal Hypertension II. Oxford, UK: Blackwell, 1996: 989.
  • 19
    Grace ND, Groszmann RJ, Garcia-Tsao G, et al. Portal hypertension and variceal bleeding: an AASLD single topic symposium. Hepatology 1998; 28: 86880.
  • 20
    de Franchis R, Merkel C, Bolondi L, et al. Committee for Portal Hypertension of the Italian Association for the Study of the Liver. AISF practice guidelines for portal hypertension. Ital J Gastroenterol Hepatol 1999; 31: 22434.
  • 21
    Jalan R, Hayes PC. UK guidelines on the management of variceal haemorrhage in cirrhotic patients (British Society of Gastroenterology). Gut 2000; 46 (Suppl. 3): III1III15.