Potential conflict of interest: Nothing to report.
Budd-Chiari syndrome (BCS) causes postsinusoidal portal hypertension, which leads to complications similar to those observed in cirrhosis. However, no studies have investigated whether patients with BCS develop the hyperdynamic circulatory syndrome present in patients with cirrhosis who have portal hypertension. We evaluated systemic and cardiopulmonary hemodynamics, plasma renin activity, aldosterone and norepinephrine levels, and plasma volume in patients with BCS admitted for complications of portal hypertension. BCS patients had mean systemic and cardiopulmonary pressures and cardiac indices that were within the normal range but were significantly different from those of a group of patients with cirrhosis matched by sex, body surface, and liver function (cardiac index 3.1 ± 0.7 vs. 4.9 ± 1.2 L · min−1 · m−2; P < .001; systemic vascular resistance [SVR] index, 2,189 ± 736 vs. 1,377 ± 422 dyne · s · cm−5 · m−2, P < .001). Despite normal systemic vascular resistance, BCS patients had activation of the neurohumoral vasoactive systems, as evidenced by increased plasma renin activity, aldosterone and norepinephrine levels (15.0 ± 21.5 ng/mL · h, 76.7 ± 106.8 ng/dL, 586 ± 868 pg/mL; respectively) and plasma volume expansion. The analysis of individual BCS patients identified that 7 of the 21 patients actually had reduced SVR index. These patients had the greatest plasma volume expansion. A significant inverse correlation between plasma volume and SVR index was observed. In conclusion, patients with BCS had activation of vasoactive neurohumoral systems and expanded plasma volume. This outcome was observed even though most of these patients did not exhibit systemic vasodilation and cardiac output was not increased, in marked contrast with what is observed in patients with cirrhosis. (HEPATOLOGY 2006;43:27–33.)
Budd-Chiari syndrome (BCS) is an uncommon liver disease defined as an obstruction to hepatic venous outflow at any level from the small hepatic veins to the junction of the inferior vena cava and the right atrium.1 The obstruction of the hepatic venous outflow leads to sinusoidal congestion, centrilobular necrosis, fibrosis, and portal hypertension. As in patients with cirrhosis, complications of portal hypertension such as variceal bleeding, refractory ascites, and hepatorenal syndrome may occur.2, 3
In cirrhosis, portal hypertension triggers the development of a marked peripheral and splanchnic arteriolar vasodilation that is thought to be the key factor promoting the activation of neurohumoral systems leading to sodium and water retention and formation of ascites, in accordance with the peripheral arterial vasodilation hypothesis.4 Furthermore, experimental and clinical studies indicate that expansion of plasma volume allows the maintenance of an increased cardiac output.5–7 This so-called “hyperdynamic circulatory syndrome” further increases portal pressure by increasing splanchnic blood flow.
Although these hemodynamic changes are well described in patients with cirrhosis who have sinusoidal portal hypertension, it is uncertain if BCS patients with postsinusoidal portal hypertension share the same hemodynamic abnormalities. The present study was performed to assess systemic circulation, vasoactive neurohumoral systems, and plasma volume in a large series of patients with advanced BCS and to compare the systemic hemodynamics of these patients with those obtained in a group of patients with cirrhosis matched by sex, body surface area, and liver function.
The study included 21 patients with severe BCS who were referred to the Hepatic Hemodynamic Laboratory of the Barcelona Liver Unit for evaluation and treatment from 1989–2005. The diagnosis of BCS was based on clinical and Doppler ultrasound and/or computed tomography scan and/or magnetic resonance imaging findings and further confirmed with hepatic venography in all cases. The diagnostic work-up included a systematic screening for any underlying prothrombotic disorder as previously described.2, 8 Analytical determinations as well as a detailed clinical exploration were performed allowing the evaluation of the degree of liver failure as assessed by Child–Pugh, Clichy, and Rotterdam scores.9–11 In addition, the presence of gastroesophageal varices was investigated via upper gastrointestinal endoscopy.
A group of 21 consecutive patients with cirrhosis admitted for evaluation of portal hypertension matched by sex, body surface area, and liver function were used for comparison. All patients gave informed consent to participate in the study after receiving complete and detailed information of the hemodynamic and laboratory investigations.
After fasting overnight, the patients were transferred to the Hepatic Hemodynamic Laboratory. Under local anesthesia, an 8-F venous catheter introducer (Axcess; Maxxim Medical, Athens, TX) was placed in the right jugular vein under ultrasonographic guidance (SonoSite Inc., Bothell, WA) using the Seldinger technique. After hepatic venography confirmed the partial or complete occlusion of the hepatic veins, a Swan-Ganz catheter (Edwards Laboratory, Los Angeles, CA) was advanced into the pulmonary artery under fluoroscopic control for measurement of cardiopulmonary pressures and cardiac output (CO) by thermal dilution. Mean arterial pressure was measured every 5 minutes with a noninvasive automatic sphygmomanometer (Marquette Electronics, Milwaukee, WI). Heart rate was derived from continuous electrocardiogram monitoring. The systemic vascular resistance (SVR) (dyne · s · cm−5) was calculated as mean arterial pressure (mm Hg) − right atrial pressure (mm Hg) × 80/CO (L · min−1). Pulmonary vascular resistance (PVR, dyne · s · cm−5) was calculated as [pulmonary artery pressure − pulmonary capillary pressure × 80/CO]. SVR and CO were indexed by body surface area and expressed as SVR index (dyne · s · cm−5 · m−2) and cardiac index (L · min−1 · m−2) respectively. No patient had tense ascites at the time of the hemodynamic evaluation or was receiving vasoactive drugs during the previous week.
Blood samples were obtained in 13 patients during the hemodynamic study for neurohumoral determinations. The samples were placed on ice, centrifuged at 4°C and stored at −70°C until analysis. Plasma renin activity, aldosterone levels, and norepinephrine levels were determined via radioimmunoassay (Clinical Assay, Baxter, Cambridge, MA; Coat-a-Count Aldosterone, Diagnostic and Products Corp., Los Angeles, CA; and IBL Laboratories, Hamburg, Germany, respectively) as previously described.12
Plasma Volume Determinations.
Plasma volume was estimated using 125I-labeled albumin as previously described13 as well as indirect measurement derived from 51Cr-labeled red blood cell volume calculation and hematocrit14 according to the standardized clinical protocol at our hospital to rule out myeloproliferative syndromes.15 In patients with ascites, body weight used in calculations was estimated in two different ways: weight estimated without ascites (dry weight) and the ideal weight based on height.16 Plasma volume was expressed in milliliters and was adjusted for weight (mL/kg). In 14 patients, plasma volume determination was performed 7.7 ± 8.1 days (range 0–28) after the hemodynamic evaluation and always before or at least 7 days after a paracenthesis with albumin reposition. In 7 patients, the hematology study was performed more than 1 month before or after the hemodynamic evaluation and was not included in the analysis. Nine healthy individuals matched by age and sex were included as controls.
Statistical analyses were performed using the SPSS version 11.0 statistical package (SPSS Inc., Chicago, IL). All results are expressed as the mean ± SD or as frequencies (%). Comparisons within each group were performed using the Student t test for paired data or the Wilcoxon test, and comparisons between groups were performed with the Student t test for unpaired data or the Mann-Whitney U test when appropriate. ANOVA followed by preplanned contrast analysis was specifically used to compare differences in plasma volume between etiological groups. Qualitative variables were compared using the Fisher's exact test. Correlation was performed using Pearson's coefficient. Significance was established at a P level of .05.
The clinical characteristics of the BCS patients are summarized in Table 1. The etiology of BCS was a myeloproliferative disorder in 38% of patients (5 polycythemia vera and 3 essential thrombocythemia) and another prothrombotic disorder in 33% (4 primary antiphospholipid syndrome, 2 paroxysmal nocturnal hemoglobinuria, and 1 Factor V Leiden mutation), and 6 patients (28%) remained idiopathic. One patient had polycythemia vera associated with Factor V Leiden mutation.
Table 1. Clinical Data of BCS and Patients With Cirrhosis Included in the Study
The mean time from the first symptoms of BCS until the hemodynamic study was 23.6 ± 53.7 months (range 0.3–205). At the time of inclusion (i.e., the hemodynamic study), 18 (86%) patients had ascites, 12 of whom were receiving diuretics. The Child–Pugh, Clichy, and Rotterdam scores at inclusion (mean 8.2 ± 1.8, 5.6 ± 0.9, and 1.3 ± 0.9, respectively) demonstrated poor liver function.
After the hemodynamic study, patients were followed up until May 2005 or death (36.2 ± 40 mo; range 0.4–137). Sixteen patients required a transjugular intrahepatic portal-systemic shunt (TIPS) (13 because of refractory ascites and 3 because of esophageal variceal bleeding; 2 of these patients required liver transplantation 24 and 26 months later, and 1 died 15 days after TIPS due to liver failure). A side-to-side portal-caval shunt was performed in 2 patients (because of variceal bleeding and refractory ascites, respectively) who died 27 months and 45 days after the surgical procedure due to pneumonia and surgery complications, respectively. In 1 additional patient, a short hepatic vein stenosis was dilated and stented. The remaining 2 patients are alive without clinical manifestations of portal hypertension, and are undergoing close follow-up (30 and 7.1 mo, respectively).
Hemodynamic and Laboratory Findings
Patients with BCS had mean systemic and pulmonary hemodynamics within the normal range (Table 2). Indeed, despite the presence of severe complications of portal hypertension and liver failure, most BCS patients did not exhibit a hyperdynamic circulation. Thus, mean SVR was not reduced and CO was not increased. These findings are contrary to those seen in patients with cirrhosis who had a similar degree of portal hypertension and liver failure that exhibited the expected hyperdynamic circulation with significantly lower SVR and significantly higher CO (Table 2). Similar results were obtained when body weight was estimated as a dry or ideal weight.
Table 2. Systemic and Pulmonary Hemodynamics in Patients With BCS and Cirrhosis
BCS (n = 21)
Cirrhosis (n = 21)
Abbreviations: CI, cardiac index; MAP, mean arterial pressure; NS, not significant; SVRI, SVR index; RAP, right atrial pressure; PAP, pulmonary arterial pressure; PCP, pulmonary capillary pressure; PVR, pulmonary vascular resistance.
In those patients requiring TIPS, the portal pressure gradient was measured directly during the procedure (54.2 ± 133 d after the initial hemodynamic study; range 0–492). All patients had severe portal hypertension (mean portal-caval pressure gradient: 24 ± 6 mm Hg). The control group of patients with cirrhosis had a similar degree of portal hypertension estimated according to the hepatic venous pressure gradient (20.7 ± 3.8 mm Hg). As shown in Table 1, patients with cirrhosis were well matched by sex, body surface area, and degree of liver failure, as evaluated using Child–Pugh scores, to BCS patients.
Patients with BCS exhibited an activation of the neurohumoral vasoactive systems, as evidenced by the increased plasma renin activity (15.0 ± 21.5 ng/mL · h; normal values in our laboratory: 1.2 ± 0.9 ng/mL · h), aldosterone (76.7 ± 106.8 ng/dL; normal values: 24 ± 2 ng/dL) and norepinephrine (586 ± 868 pg/mL; normal values: 233 ± 17 pg/mL) (Fig. 1). Activation of the renin-angiotensin-aldosterone system was associated with plasma volume expansion (Table 3). Similar results were observed independently of the method used to determine blood volume (Table 3). The plasma volume was expanded to a similar degree independently of BCS etiology (myeloproliferative disorder, 49.7 ± 16.7 mL/kg; other prothrombotic disorder, 49.3 ± 23.6 mL/kg; idiopathic, 53.4 ± 19.3 mL/kg, P > .1).
Table 3. Plasma Volume in BCS Patients and Healthy Controls
BCS (n = 14)
Healthy Controls (n = 9)
Abbreviations: PV, plasma volume; 125I, 125I-labeled albumin; 51Cr, 51Cr-labeled red blood cells; NS, not significant.
Twelve patients with ascites had initiated diuretic treatment (100 mg spironolactone, 40 mg furosemide) when the hemodynamic/neurohumoral/plasma volume evaluation was performed. However, among patients receiving or not receiving diuretics, there were no significant differences in systemic and cardiopulmonary pressures (Table 4), degree of plasma volume expansion (51.0 ± 21.7 mL/kg vs. 50.5 ± 14.6 mL/kg; P value not significant), or activation of the neurohumoral systems (Fig. 1).
Table 4. Hemodynamic Results in BCS Patients With and Without Diuretics
With Diuretics (n = 12)
Without Diuretics (n = 9)
Abbreviations: NS, not significant; MAP, mean arterial pressure; RAP, right atrial pressure; PAP, pulmonary arterial pressure; PCP, pulmonary capillary pressure; PVR, pulmonary vascular resistance.
CO (L · min−1)
5.5 ± 1.5
5.4 ± 1.4
MAP (mm Hg)
86 ± 13.6
85 ± 10.7
SVR (dyne · s · cm−5)
1,283 ± 436
1,297 ± 565
RAP (mm Hg)
4.4 ± 2.7
4.8 ± 4.3
PAP (mm Hg)
12.9 ± 6.1
12 ± 4.5
PCP (mm Hg)
6.5 ± 2.7
6.5 ± 4.1
PVR (dyne · s · cm−5)
90.7 ± 57.7
94.2 ± 62.8
Analyses of individual BCS patients identified that 7 of the 21 patients actually had a low SVR index (below normal values: mean 1,412 ± 193 dyne · s · cm−5 · m−2; range 1,168–1,594 dyne · s · cm−5 · m−2). These patients had significantly lower mean arterial pressure and a higher cardiac index, although the latter was still within normal values (Table 5) compared with nonvasodilated BCS patients. In addition, vasodilated BCS patients had greater plasma volume expansion (63.5 ± 23.7 mL/kg vs. 43.9 ± 10.2 mL/kg in nonvasodilated patients; P = .1). However, BCS patients with nonreduced SVR still exhibited plasma volume expansion compared with healthy controls (43.9 ± 10.2 mL/kg vs. 34.8 ± 5.8 mL/kg; P = .03). In the overall series of BCS patients, plasma volume had a highly significant inverse correlation with SVR (r = −0.90, n = 14; P < .001) and SVR index (r = −0.89, n = 14; P < .001) (Fig. 2). No significant differences in sex, age, or clinical symptoms (ascites, varices, and prognostic score indexes) were observed between patients with or without reduced SVR (Table 5). However, in patients with reduced SVR, the clinical presentation of BCS was more frequently chronic (7/7 in reduced SVR vs. 4/14 in nonreduced SVR; P = .004).
Table 5. Clinical Characteristics and Hemodynamics of Patients With Vasodilated and Nonvasodilated BCS
Vasodilated (n = 7)
Nonvasodilated (n = 14)
Abbreviations: NS, not significant; M, male; F, female; CI, cardiac index; SVRI, SVR index; MAP, mean arterial pressure.
Hematocrit was not significantly different between BCS patients and healthy controls (Table 3). However, in patients with cirrhosis, the hematocrit was significantly lower compared with BCS patients (32 ± 5% vs. 40 ± 8%; P < .01). Hematocrit, but not total red cell mass, significantly correlated with cardiac index (r = −0.50, n = 21; P = .01) (Fig. 3) in BCS patients but not in patients with cirrhosis.
In 4 patients, a second cardiopulmonary evaluation was performed after TIPS (immediately in 2 cases, and 1 d and 3 mo after in the remaining 2 cases). After TIPS, CO increased in all cases by a mean of 60%.
This study evaluated the systemic hemodynamics, vasoactive neurohumoral systems, and plasma volume in a large series of patients with BCS and severe portal hypertension. The results show that despite the heterogeneity of patients with BCS and severe portal hypertension, the hemodynamic profile of BCS patients is markedly different from that found in patients with cirrhosis5, 6 and prehepatic portal hypertension.17, 18 Indeed, whereas patients with cirrhosis exhibited the expected hyperdynamic circulatory syndrome characterized by increased CO and reduced SVR, patients with BCS had normal cardiopulmonary hemodynamics, and most of them did not exhibit systemic vasodilation. Normal CO and SVR was also observed in a small series of a specific subset of BCS patients with associated thrombosis of the inferior vena cava.19
In patients with cirrhosis, peripheral vasodilation (mainly at the splanchnic bed) by way of increasing vascular capacity is thought to be the main trigger promoting the overactivation of vasoactive neurohumoral systems that lead to sodium retention and plasma volume expansion. However, most BCS patients exhibited a marked activation of neurohumoral vasoactive systems despite the lack of vasodilation. Vasoactive systems were also activated in patients not receiving diuretics, showing that in BCS patients these systems are overactivated independently of the administration of these drugs.
Although SVR was not reduced in the overall series, a moderate reduction in SVR was observed in one third of BCS patients. All of these patients had chronic BCS, and the study was performed late in the disease evolution, suggesting that peripheral vasodilation in BCS, if present, is a late event. These BCS patients with moderate vasodilation had greater plasma volume expansion and lower mean arterial pressure. Actually, in the overall BCS population, there was a close and inverse correlation between SVR and plasma volume, suggesting that systemic vasodilation still plays a role in modulating plasma volume expansion in this condition.
It is interesting to note that in spite of marked portal hypertension and development of portal-systemic collaterals and ascites, vasodilation was only observed in a small proportion of BCS patients. The findings of our study do not provide an explanation for the lack of systemic vasodilation in most of our BCS patients. However, we can speculate that the intense acute–subacute obstruction to hepatic blood flow may promote some vasoconstrictive stimuli20 aimed to reduce splanchnic blood flow that may, in part, compensate the vasodilatory factors triggered by portal hypertension. In addition, our BCS patients showed a mean normal hematocrit substantially higher than that seen in patients with cirrhosis. It is possible, therefore, that maintenance of blood viscosity may also prevent a further reduction in SVR.21
Activation of vasoactive systems in the absence of systemic vasodilation in a high proportion of our patients is intriguing. This may be due in part to the so-called “hepatorenal reflex,” elicited by an increase in intrahepatic pressure that could promote the activation of the renin-angiotensin-aldosterone axis.22 In fact, early studies conducted by Orloff et al.23 showed that dogs with experimentally induced minimal hepatic venous occlusion exhibited a marked increment in aldosterone secretion before any change in plasma volume, blood pressure, or systemic hemodynamics. Similarly, a reduction in renal blood flow and glomerular filtration rate and reduced sodium excretion have been found in experimental models of portal hypertension secondary to hepatic venous outflow obstruction24, 25 and in patients with cirrhosis in whom TIPS was transiently occluded.26 In addition, in patients with cirrhosis and portal hypertension, a reduced venous tone with increased venous compliance and capacitance has been shown to be an important component contributing to the effective hypovolemic state that further activates the neurohumoral systems.27–29 Therefore, it is conceivable that the splanchnic venous bed, the portal-systemic collaterals, and the systemic venous circulation might contribute to reduce the effective blood volume in BCS, especially in the early phases of the disease, further triggering the activation of the neurohumoral systems and plasma volume expansion.
Contrary to patients with cirrhosis, BCS patients do not have increased CO. In cirrhosis, once the permissive factor of expanded plasma volume is present, CO increases. This is believed to be a compensatory mechanism to maintain arterial pressure from systemic vasodilation; additionally, it is believed to preserve an adequate tissue oxygen supply in spite of the frequent anemia found in cirrhosis.21 In that regard, we have previously shown that an increase in the low blood hemoglobin levels decreases CO in patients with cirrhosis.30 The absence of anemia in BCS patients may explain at least in part the lack of increase in CO. This is supported by our finding of an inverse correlation between hematocrit and the cardiac index. Alternatively, it may be suggested that the lack of increase in CO—despite plasma and blood volume expansion—might be explained by reduced venous return from the inferior vena cava,31, 32 because of the compression by a hypertrophied caudate lobe,33 or from the splanchnic vascular bed because of outflow obstruction.26 The marked increase in CO observed after increasing venous return by TIPS placement clearly shows that the heart of BCS patients is able to have this adaptative response.
In summary, patients with severe BCS had activation of vasoactive neurohumoral systems and expanded plasma volume. The systemic hemodynamics in patients with BCS are heterogeneous. In some patients, usually those with longer duration of the disease, these abnormalities are associated with peripheral vasodilation. However, two thirds of patients did not exhibit systemic vasodilation despite plasma volume expansion and activation of vasoactive systems, which is in marked contrast with what is observed in patients with cirrhosis.
The authors are indebted to M. A. Baringo, L. Rocabert, R. Saez, and N. Campos for their expert technical assistance and M. Montaño for editorial support.