LIVER FAILURE AND LIVER DISEASE
Nadolol Plus Spironolactone in the Prophylaxis of First
Variceal Bleed in Nonascitic Cirrhotic Patients:
A Preliminary Study
Raquel Abecasis,1David Kravetz,1Eduardo Fassio,2Beatriz Ameigeiras,3Daniel Garcia,4Rogelio Isla,5
Graciela Landeira,2Nora Dominguez,2Gustavo Romero,1Julio Argonz,1and Ruben Terg1
Treatment with ?-blockers fails to decrease portal pressure in nearly 40% of cirrhotic patients.
Recent studies have suggested that treatment with spironolactone reduces pressure and flow in
more effective than nadolol alone to prevent the first variceal bleeding. One hundred patients
with medium and large varices who had never bled and were without ascites were included in a
randomized into 2 groups: 51 received nadolol plus placebo (N ? P) and 49 received nadolol
24 patients. There were no significant differences in the appearance of variceal bleeding and
ascites between groups at a mean follow-up of 22 ? 16 months. However, analyzing both
complications together, the incidence was significantly higher in the N ? P group than in the
N ? S group (39% vs. 20%; P < .04). Clinical ascites was also higher in patients in the N ? P
group than in the N ? S group (21% vs. 6%; P < .04). Significant increases in plasma renin
activity and plasma aldosterone levels were only observed in patients in the N ? S group (P <
groups after 70 months of follow-up. In conclusion, these results suggest that nadolol plus
spironolactone does not increase the efficacy of nadolol alone in the prophylaxis of the first
variceal bleeding. However, when bleeding and ascites were considered together, the combined
therapy effectively reduced the incidence of both portal-hypertensive complications.
large esophageal varices.1Previous studies have shown
that reduction of the hepatic venous pressure gradient
t present, nonselective ?-adrenergic blockers (?-
blockers) are the drugs of choice to prevent the
first variceal bleeding in cirrhotic patients with
(HVPG) to less than 12 mm Hg or a decrease in HVPG
greater than 20% from basal values protects against
variceal hemorrhage. However, such a decrease in portal
ing ?-blockers.2Moreover, 40% of treated patients do
not have reduced portal pressure despite adequate
portal pressure in a greater proportion of patients.4
Recent data have shown that spironolactone signifi-
cantly lowers portal and variceal pressures by reducing
plasma volume and splanchnic blood flow.5-9Plasma vol-
ume depletion improves the hyperdynamic circulatory
portal hypertension.10,11However, a correlation could
not be shown between the decrease in circulating plasma
volume and the decrease in HVPG.5,6,8It has been sug-
gested that spironolactone may also have a direct vasoac-
tive effect on the splanchnic circulation that is not
mediated by its antialdosteronic mechanism.12
Abbreviations: HVPG, hepatic venous pressure gradient; N ? P, nadolol plus
placebo; N ? S, nadolol plus spironolactone.
From the1Liver Unit, Hospital de Gastroenterologı ´a B. Udaondo;2Department
of Gastroenterology, Hospital A. Posadas;3Department of Gastroenterology, Hospi-
tal Ramos Mejia;4Department of Gastroenterology, Hospital Municipal; and5De-
partment of Gastroenterology, Hospital Israelita, Buenos Aires, Argentina.
Received June 11, 2002; accepted October 30, 2002.
Supported by a grant from the Fundacio ´n Argentina para el Estudio de las
Enfermedades del Hı ´gado.
This study was presented as an oral presentation at the 2001 annual meeting of
the American Gastroenterological Association.
Address reprint requests to: Raquel Abecasis, M.D., 25 de Mayo 229, Bernal,
Buenos Aires, Argentina (cp:1876). E-mail: email@example.com;
fax: (54) 11-4306-2033.
Copyright © 2003 by the American Association for the Study of Liver Diseases.
The decrease in portal pressure achieved by adminis-
tration of spironolactone could be found even in patients
diet.6-8The acute addition of intravenous propranolol to
long-term spironolactone therapy (100 mg/d) further de-
creased HVPG.7Furthermore, the combination of spi-
ronolactone and propranolol significantly reduced
variceal pressure in the subset of patients unresponsive to
However, whether these hemodynamic effects have
any impact on the management of cirrhotic patients with
portal hypertension has never been assessed in long-term
clinical trials. Therefore, the aim of this study was to
compare the nonselective ?-blocker nadolol with spi-
ronolactone plus nadolol in the primary prophylaxis of
variceal bleeding. We also investigated whether the com-
bined treatment would have a beneficial effect to prevent
the first appearance or recurrence of ascites.
Patients and Methods
Patients. From October 1993 to December 1999,
100 patients seen in 5 centers in Argentina were enrolled:
45 in Hospital B. Udaondo, 36 in Hospital Posadas, 8 in
Hospital Ramos Mejı ´a, 6 in Hospital Israelita, and 5 in
Hospital Municipal. The inclusion criteria were as fol-
lows: (1) a diagnosis of cirrhosis based on clinical and
the procedure was not contraindicated; (2) medium and
large esophageal varices irrespective of the presence of red
color signs13; (3) no history of variceal bleeding; (4) ultra-
sonographic absence of ascites; and (5) no diuretic treat-
exclusion criteria were as follows: (1) contraindications to
administration of ?-blockers (heart rate ?55 beats/min,
systolic blood pressure ?85 mm Hg, chronic obstructive
lung disease, psychosis, insulin-dependent diabetes with
history of hypoglycemia, heart failure, and second- or
third-degree atrioventricular blocks); (2) hepatocellular
carcinoma or other malignancies; (3) serum urea and/or
creatinine levels greater than 50 and 1.5 mg/dL, respec-
tively; (4) bacterial infections; and (5) chronic hepatic
encephalopathy. According to these criteria, 100 of 170
consecutive patients were eligible for this study. Reasons
for exclusion were as follows: small varices (n ? 10), pre-
vious variceal bleeding (n ? 15), ascites (n ? 11), small
ers (n ? 12), chronic encephalopathy (n ? 3), tumors
(n ? 6), and refusal to participate in the study (n ? 3).
The study was designed as a randomized, multicenter,
prospective, double-blinded trial. The protocol, which
conformed with the Helsinki Declaration, was approved
by the hospital ethical committee of each center, and pa-
tients gave their written informed consent to participate
in the investigation. Randomization was performed by
tables of random numbers at the coordinating center and
stratified according to participating hospitals.
Protocol. After randomization, patients were treated
with nadolol at increasing doses to reduce resting heart
rate by 25% or to 55 beats/min. Once the suitable dose
P) received a placebo tablet and 49 patients treated with
nadolol plus spironolactone (N ? S) received spironolac-
tone at a fixed dosage of 100 mg/d for the whole study.
Spironolactone and placebo, in identical tablets, were
given by an assistant not directly involved with the pro-
tocol to ensure study blindness. Treatment was adminis-
tered once daily at breakfast. All patients were following
an unrestricted-sodium diet. Compliance with treatment
was assessed by checking persistence of heart rate reduc-
tion and counting the remaining tablets at each visit. Pa-
tients who consumed less than 85% of prescribed pills
were classified as noncompliant.
Patients were followed up monthly with a physical ex-
amination as well as measurement of heart rate and arte-
rial blood pressure and every 2 and 6 months with
biochemical evaluation and abdominal ultrasonography,
respectively, for the first 2 years. Thereafter, the clinical
evaluation was performed every 3 months and the bio-
until end point or December 1999.
each group before and after 2 to 3 months of treatment.
After an overnight fast and under local anesthesia, a 7F
vein by the Seldinger technique. Under fluoroscopy, a 7F
balloon-tipped catheter (Medi Tech; Cooper Scientific
Corp., Watertown, MA) was advanced into the main
right hepatic vein. Wedge (occluded) and free hepatic
the balloon. Afterward, a 7F Swan-Ganz catheter (Ed-
wards Laboratory, Los Angeles, CA) was placed into the
pulmonary artery to measure cardiopulmonary pressures
and cardiac output (thermodilution). Arterial pressure
was recorded by a sphygmomanometer attached to the
right arm of the patient, who was in the supine position,
and heart rate was derived from the continuous electro-
cardiographic monitoring. All parameters were measured
at least in triplicate, and tracings were obtained on a mul-
tichannel recorder (Electronics Inc., Pleasantville, NY).
ence between wedge hepatic venous pressure and free he-
s ? cm?5) was calculated as (MAP ? right atrial pressure)/
360ABECASIS ET AL.HEPATOLOGY, February 2003
CO ? 80 in which MAP indicates mean arterial pressure
(mm Hg) and CO indicates cardiac output (L/min).14
After an overnight fast and 2 hours of bed rest, blood
samples from 12 patients in each group were taken to
els. Samples were collected in tubes containing ethyl-
enediaminetetraacetic acid that were placed on ice and
centrifuged at 4°C, and the plasma was frozen at ?30°C
ng/mL/h) and plasma aldosterone levels (normal value,
35-350 pg/mL) were measured by radioimmunoassay be-
fore and after 18 months of treatment.
The primary end point of the trial was variceal bleed-
ing. Secondary end points were appearance of ascites, ad-
verse effects requiring withdrawal from treatment, and
survival. We diagnosed variceal bleeding when (1) a pa-
tient with hematemesis and/or melena showed varices ac-
(2) varices without another potential source of bleeding
were identified at emergency endoscopy performed
within the first 24 hours of the hemorrhage.1These pa-
tients were treated with sclerotherapy and were with-
drawn from the trial. Patients who developed either
minimal ascites, only detectable by ultrasonography, or
than 85 mm Hg; heart failure; encephalopathy without
improvement despite a low-protein diet and lactulose
therapy; bronchospasm; arrhythmia; severe painful gy-
necomastia; serum urea and/or creatinine levels greater
than 50 and 1.5 mg/dL, respectively; hyponatremia (a
decrease in serum sodium ?5 mEq/L to a level ?130
mEq/L); or hyperkalemia (an increase ?1.5 mEq/L to a
level ?5 mEq/L).
Statistical Analysis. Taking into account that no
long-term clinical data are available regarding the effects
of spironolactone on portal pressure and that patients
with large esophageal varices treated with ?-blockers
would have an expected risk of bleeding of 31% during a
follow-up of at least 2 years, we considered it clinically
relevant to decrease this risk to 12%. In that regard, we
0.05 and a ? error of 0.2 in a 2-tailed test. An interim
analysis was planned when most patients completed at
least 2 years of follow-up.
Comparisons between groups were performed using
nested ANOVA test for quantitative variables and ?2test
for qualitative variables. The cumulative probabilities of
variceal bleeding, ascites, and survival were calculated us-
ing the Kaplan-Meier method and compared by log-rank
test. Predictors of bleeding were identified by Cox’s re-
gression model. The following variables were assessed:
type of treatment, age, sex, cause of cirrhosis, Child-Pugh
than .05 were considered significant. Results are given as
mean ? SD. Data were analyzed according to the inten-
There were no significant differences in demographic,
etiologic, clinical, and endoscopic characteristics; previ-
ous episodes of ascites; mean dosage of nadolol (78 ? 61
mg/d); and mean follow-up (22 ? 16 months) between
both groups of patients (Table 1).
to follow-up: 4 patients in the first 6 months and 2 pa-
tients at 12 and 18 months, respectively.
Five patients in the N ? P group and 4 patients in the
N ? S group had to be withdrawn from treatment be-
cause of adverse effects of ?-blockers: 3 because of brady-
arrhythmia, 3 because of bronchospasm, one because of
symptomatic hypotension, one because of encephalopa-
thy, and one because of postural dizziness and severe as-
thenia. Five patients were withdrawn during the first 6
months of treatment, and the remaining 4 were with-
drawn after 24 months of follow-up. Two patients with
mmol/L] and serum creatinine level of 1.7 mg/dL) were
withdrawn at 6 and 15 months of treatment. These com-
plications disappeared after discontinuation of the rele-
vant drug. Minor adverse effects that did not preclude
continuation with treatment included mild asthenia in 9
patients, gynecomastia in 4 patients (2 in each group),
dizziness in 3 patients, arterial hypotension in 7 patients,
and encephalopathy stage I in 3 patients, with similar
distribution in both groups (Table 2).
Inadequate compliance was observed in 4 patients in
the N ? P group and 5 patients in the N ? S group.
Table 1. General Characteristics of Included Patients
N ? P
(n ? 51)
N ? S
(n ? 49)
Alcohol consumption (%)
Child-Pugh class (A/B/C)
Esophageal varices (GII/III)
Previous ascites (n)
Mean dosage of nadolol (mg/d)
Mean follow-up (mo)
56 ? 10
5.8 ? 1.0
80 ? 60
21 ? 15
59 ? 11
6.0 ? 1.0
76 ? 62*
23 ? 18†
NOTE. No significant difference was observed between groups.
*Range of 10-240 mg/d in both groups.
†Range of 3-60 months in the N ? P group and 1-70 months in the N ? S
HEPATOLOGY, Vol. 37, No. 2, 2003ABECASIS ET AL.361
Hepatic and renal function test results and electrolyte
levels were similar before and after treatment in both
groups. Treatment with nadolol alone did not produce
any changes in plasma renin activity and plasma aldoste-
rone levels. In contrast, the addition of spironolactone to
nadolol was associated with significant increases in both
parameters (Table 3).
group, HVPG decreased significantly from 16.3 ? 3.7
mm Hg at baseline to 14.3 ? 3.5 mm Hg after treatment
(mean decrease, ?11.5 ? 14.4%; P ? .05). Patients in
from 17.9 ? 2.7 to 15 ? 3.3 mm Hg (mean decrease,
?16 ? 12.4%; P ? .01). No significant differences were
observed when the decrease in HVPG was compared be-
tween groups (Table 4).
Systemic Hemodynamics. Both groups of patients
showed effective ?-blockade, evidenced by a significant
reduction in cardiac output and heart rate in addition to
an increase in systemic vascular resistance. A small but
in the N ? S group (Table 3). There were no changes in
cardiopulmonary pressures in both groups of patients.
Clinical Outcome. During follow-up, 7 patients in
the N ? P group (13.7%) and 3 patients in the N ? S
group (6%) had variceal bleeding, Eight patients bled
from esophageal varices (6 in the N ? P group and 2 in
the N ? S group), either on medium or large varices.
These patients were treated with sclerotherapy. The re-
maining 2 patients (one in each group) bled from gastric
varices, and one of them received a portocaval shunt. The
cumulative probabilities of patients to be free of bleeding
after 70 months of follow-up were 73% in the N ? P
group and 79% in the N ? S group (P ? NS) (Fig. 1).
Ascites developed in 13 patients in the N ? P group
(25%) and in 7 patients in the N ? S group (14%).
Although the incidence of ascites was similar between
groups, we observed that 11 patients in the N ? P group
(21%) and 3 in the N ? S group (6%) had clinical ascites
(P ? .04); the number of remaining patients with mini-
mal ascites was similar in both groups (P ? NS). Ascites
de novo appeared in 10 of 13 patients in the N ? P group
and in 5 of 7 patients in the N ? S group (P ? NS). Of
the 19 patients with previous ascites, only 5 had a recur-
be free of ascites after 60 months of follow-up were 57%
in the N ? P group and 73% in the N ? S group (P ?
NS) (Fig. 2).
Considering variceal bleeding and ascites together in
complications (P ? .04).
Type of treatment, dosage of nadolol, age, sex, Child-
Pugh score, cause of cirrhosis, and previous episodes of
ascites did not have an independent predictive value of
Table 2. Side Effects During the Study
N ? P
(n ? 51)
N ? S
(n ? 49)
Patients requiring withdrawal
NOTE. No significant difference was observed between groups.
Table 3. Biochemical Parameters and Endogenous Vasoactive System Before and After Both Treatments
N ? P (n ? 51)N ? S (n ? 49)
Serum albumin (g/dL)
Prothrombin index (%)
Serum bilirubin (mg/dL)
Serum creatinine (mg/dL)
Serum sodium (mEq/d)
Urinary sodium (mEq/d)
Plasma renin activity (ng/ml/h)*
Plasma aldosterone (pg/dL)*
39 ? 5
3.5 ? 0.6
72 ? 17
1.4 ? 1.1
0.8 ? 0.1
140 ? 5
122 ? 75
1.8 ? 1.7
87 ? 59
38 ? 5
3.3 ? 0.5
72 ? 18
1.5 ? 1.2
0.8 ? 0.2
140 ? 4
114 ? 82
1.6 ? 1.5
83 ? 48
41 ? 5
3.6 ? 0.6
76 ? 18
1.3 ? 0.7
0.9 ? 0.2
139 ? 5
118 ? 61
1.6 ? 1.3
79 ? 55
40 ? 7
3.5 ? 0.6
80 ? 15
1.2 ? 0.8
0.9 ? 0.2
136 ? 7
122 ? 58
4.4 ? 2.4†
268 ? 216†
Abbreviations: B, before both treatments; A, after both treatments.
*Only measured in 12 patients in each group before and after 18 months of treatment.
†P ? .01.
362ABECASIS ET AL.HEPATOLOGY, February 2003
bleeding when tested by Cox’s multiple regression analy-
Three patients in the N ? P group and one patient in
the N ? S group died. A patient in the N ? P group died
of variceal bleeding, and the other 2 died of lung and
bladder cancer, respectively. The cause of death in the
patient in the N ? S group was hepatic failure. The cu-
mulative probability of survival was 91% in the N ? P
group and 97% in the N ? S group (Fig. 3).
When the results were analyzed at a mean follow-up of
different between both groups of patients. Considering
that the cumulative risk of bleeding and ascites was 27%
and 43%, respectively, in the N ? P group and 21% and
27% in the N ? S group, we estimated that a sample size
of 534 and 432 patients, respectively, would be necessary
to achieve a statistical difference between both groups
(? ? 0.05 and ? ? 0.20). Therefore, we decided to stop
the trial because we would be unable to recruit such a
number of patients.
Hemodynamic studies have shown that spironolactone
may be effective in decreasing portal and variceal pres-
spironolactone to protect cirrhotic patients with esophageal
varices at risk for bleeding has not yet been investigated.
We report the results of the first long-term, prospec-
tive, double-blind, randomized clinical study comparing
the efficacy and safety of the association of nadolol and
spironolactone with nadolol alone in the prophylaxis of
first variceal bleeding in nonascitic cirrhotic patients.
Taking into account that a group of patients included
in this study would be treated with spironolactone, we
also examined its ability to prevent the development of
ascites as a secondary end point.
Previous hemodynamic studies have shown that treat-
ment with spironolactone decreased portal and variceal
pressures in human and experimental cirrhosis.5-9,15,16
Moreover, the acute intravenous addition of propranolol
Fig. 1. Cumulative probability of being free of bleeding in the N ? P
and N ? S treatment groups (NS).
Fig. 2. Cumulative probability of being free of ascites in the N ? P
and N ? S treatment groups (NS).
Table 4. Hemodynamic Effects of Nadolol or Nadolol Plus Spironolactone
N ? P (n ? 12)N ? S (n ? 12)
Basal 2 MoBasal 2 Mo
MAP (mm Hg)
RAP (mm Hg)
PAP (mm Hg)
PCP (mm Hg)
SVR (dyne ? s ? cm5)
HVPG (mm Hg)
WHVP (mm Hg)
FHVP (mm Hg)
75.3 ? 19
87 ? 9
3.3 ? 1.7
12 ? 4.6
7.8 ? 3
7.3 ? 1.2
930 ? 166
16.3 ? 3.7
21.8 ? 5.7
5.6 ? 3
56 ? 3*
85 ? 8
4.2 ? 1.9
14 ? 3.7
8 ? 2.6
5.7 ? 1.4*
1,197 ? 261*
14.3 ? 3.5*
21.3 ? 3.8
7 ? 1.2
76 ? 11
94 ? 8
4.9 ? 2.1
13 ? 3.3
7.5 ? 2.8
7 ? 2.6
1,067 ? 358
17.9 ? 2.1
23.8 ? 3.3
6 ? 1.6
56 ? 4*
83 ? 8*
4.2 ? 2.9
12.6 ? 4
8.5 ? 3.1
5 ? 2.1*
1,388 ? 537
15 ? 3.3*
21.4 ? 4.7*
6.4 ? 3.3
NOTE. Four patients in the N ? P group and 5 patients in the N ? S group had a good HVPG response, without differences between groups.
Abbreviations: HR, heart rate; MAP, mean arterial pressure; RAP, right atrial pressure; PAP, pulmonary artery pressure; PCP, pulmonary capillary pressure; CO, cardiac
output; SVR, systemic vascular resistance; WHVP, wedge hepatic venous pressure; FHVP, free hepatic venous pressure.
*P ? .05 vs. baseline.
HEPATOLOGY, Vol. 37, No. 2, 2003ABECASIS ET AL. 363
to long-term administration of spironolactone further re-
duces HVPG.7Nevens et al. showed that spironolactone
also reduced variceal pressure in patients on long-term
?-blocker therapy.9On the other hand, Sugano et al. did
not find differences in the mean reduction of HVPG be-
tween patients treated with low-dose transdermal nitro-
glycerin alone or associated with spironolactone.17In the
present trial, long-term administration of nadolol plus
spironolactone produced a higher but not significant de-
crease in HVPG compared with nadolol alone (16% vs.
11.5%). This effect could be related to a decrease in ef-
fective intravascular volume suggested by the significant
increases in plasma renin activity and plasma aldosterone
levels observed in patients treated with the combined
therapy. Another possible explanation for the reduction
in HVPG could be a decrease in the intrahepatic vascular
resistance produced by the calcium channel blocking ac-
tion of spironolactone.12,18This probable vasodilator
or long-term administration of furosemide did not pro-
duce any significant changes in HVPG, total blood vol-
ume, or azygos blood flow despite a decrease in cardiac
that our patients followed an unrestricted-sodium diet.
We decided on this diet by taking into account the study
design, which included patients without ascites and their
poor compliance with a low-sodium diet for a long time.
It had been shown in portal-hypertensive rats20and in
cirrhotic patients6that restriction of sodium produces a
significant reduction in plasma volume and improves the
hyperdynamic circulation associated with cirrhosis. In
this trial, treatment with spironolactone probably did not
result in additional effects because the administered
amount of sodium blunted a more pronounced decrease
of plasma volume. However, it has also been shown that
spironolactone decreases portal and variceal pressures in
believe it is unlikely that intake of sodium plays a major
role in our results.
We observed a low incidence of bleeding in the whole
plus nadolol. In absolute terms, only 10 patients bled and
decreased almost by half in the combined treatment
group (7 vs. 3). This could be related to the inclusion of
patients with less severe liver disease (mean Child-Pugh
been shown that ascites is an important prognostic factor
increasing the risk of bleeding21and was the most impor-
tant indicator of both bleeding and death risks in a trial
comparing patients with and without ascites treated with
propranolol.22Vorobioff et al. observed that, after 3
months of treatment with propranolol, portal pressure
decreased significantly more often in patients without as-
a pathophysiologic study reporting that the density of
?-adrenoreceptors in circulating mononuclear cells is re-
duced in patients with ascites, probably resulting in a low
hemodynamic response to ?-blockers.24This evidence
suggests that ?-blockers are more effective in reducing
portal hypertension in compensated cirrhotic patients,
even more when administered at early stages before the
appearance of esophageal varices.25
Although the development of ascites was similar in
both groups, the subset of patients with clinical evidence
of this complication was significantly higher in the group
treated with nadolol alone than in those who received the
in patients who received combined treatment compared
with those treated with nadolol alone (12% vs. 25%);
however, this difference was not significant. The inci-
dence of ascites de novo in patients treated with nadolol
alone was similar to that reported in long-term trials on
primary prophylaxis of variceal bleeding with ?-blockers
ranging from 27% to 32%.22,26,27All of these findings
show that a trend toward the association of nadolol and
spironolactone may be useful in the prevention of clinical
ascites that should be treated with sodium restriction and
diuretics, whereas minimal ascites, only detectable by ul-
trasonography, does not necessitate specific treatment.28
A further analysis, according to the occurrence of
variceal bleeding and ascites in each group of patients,
showed that the incidence of both complications was sig-
nificantly higher in patients treated with nadolol alone
than in those treated with spironolactone plus nadolol.
This interesting finding supports the concept that the
Fig. 3. Cumulative probability of survival in the N ? P and N ? S
treatment groups (NS).
364ABECASIS ET AL.HEPATOLOGY, February 2003
prevention of ascites would increase the efficacy of nado- Download full-text
Although withdrawals due to adverse effects were sim-
ilar in both groups, most were related to ?-blockade
rather than to spironolactone (9 vs. 2). Only 2 patients in
the combined treatment group developed an increase in
serum creatinine and urea levels associated, in one of the
patients, with hyperkalemia that improved with discon-
tinuation of the diuretic drug. Despite the concern that
receiving spironolactone therapy,9none of the patients in
this trial had to be withdrawn for this adverse event.
In conclusion, these preliminary results suggest that
nadolol plus spironolactone is not more effective than
nadolol alone in preventing the first variceal bleeding in
nonascitic cirrhotic patients. However, this combined
treatment would prevent both complications of portal
hypertension, particularly the appearance of clinical as-
cites. Future clinical trials with a greater number of pa-
in the prophylaxis of the first variceal bleeding.
for her assistance with this study.
The authors thank Alicia Podesta ´
1. De Franchis R. Portal hypertension. Baveno III Consensus Statements.
2. Groszmann RJ, Bosch J, Grace N, Conn HO, Garcı ´a-Tsao G, Navasa M,
Alberts J, et al. Hemodynamic events in a prospective randomized trial of
propranolol vs placebo in the prevention of the first variceal hemorrhage.
3. Feu F, Garcı ´a-Paga ´n JC, Bosch J, Luca A, Tere ´s J, Escorcell A, Rodı ´s J.
Relation between portal pressure response to pharmacotherapy and risk of
recurrent variceal hemorrhage in patients with cirrhosis. Lancet 1995;346:
4. Bosch J, Garcı ´a-Paga ´n JC, Feu F, Luca A. New approaches in the pharma-
cologic treatment of portal hypertension. J Hepatol 1993;17(Suppl 2):
5. Okumura H, Arakami T, Katsuta Y, Satomura K, Akaike M, Sekiyama T,
Terada H, et al. Reduction in hepatic venous pressure gradient as a conse-
quence of volume contraction due to chronic administration of spirono-
lactone in patients with cirrhosis and no ascites. Am J Gastroenterol 1991;
6. Garcı ´a-Paga ´nJ,SalmeronJ,FeuF,GinesP,PizcuetaP,ClariaJ,ArroyoV,
et al. Effects of low-sodium diet and spironolactone on portal pressure in
patients with compensated cirrhosis. HEPATOLOGY 1994;19:1095-1099.
7. Garcı ´a-Paga ´nJ,SalmeronJ,FeuF,GinesP,PizcuetaP,ClariaJ,ArroyoV,
et al. Plasma volume reduction by spironolactone decreases portal pressure
in patients with compensated cirrhosis. Synergistic effect with propranolol
[Abstract]. J Hepatol 1991;13:530.
8. Sugano S, Kawafune T, Okajima T, Ishii K, Watanabe M, Takamura N.
Chronic splanchnic hemodynamic effects of spironolactone with unre-
stricted sodium diet in patients with compensated cirrhosis. Dig Dis Sci
9. Nevens F, Lijnen P, VanBilloen H, Fevery J. The effects of long-term
treatment with spironolactone on variceal pressure in patients with portal
hypertension without ascites. HEPATOLOGY 1996;23:1047-1052.
10. Colombato L, Albillos A, Groszmann R. Temporal relationship of periph-
eral vasodilatation, plasma volume expansion and the hyperdynamic cir-
culatory state in portal hypertension rats. HEPATOLOGY 1992;15:323-328.
11. Schrier R, Arroyo V, Bernardi M, Epstein M, Henriksen J, Rodes J. Pe-
ripheral arterial vasodilation hypothesis: a proposal for the initiation of
renal sodium and water retention in cirrhosis. HEPATOLOGY 1988;8:1151-
12. Dacquet C, Loirand G, Mironneau C, Mironneau J, Pacaud P. Spirono-
lactone inhibition of contraction and calcium channels in rat portal vein.
Br J Pharmacol 1987;92:535-544.
13. Japanese Research Society for Portal Hypertension. The general rules for
recording endoscopic findings on esophageal varices. Jpn J Surg 1980;10:
14. Romero G, Kravetz D, Argonz J, Bildozola M, Suarez A, Terg R. Terlip-
ressin is more effective in decreasing variceal pressure than portal pressure
in cirrhotic patients. J Hepatol 2000;32:419-425.
15. Van De Casteele M, Van Roel G, Nevens F, Fevery J. Effects of varying
doses of spironolactone without and with nitrates on portal vein pressure
and kidney function in partial portal vein ligated rats. HEPATOLOGY 1996;
16. Oberti F, Pilette C, Rifflet H, Maiga MY, Moreau A, Gallois Y, Girault A,
et al. Effects of simvastatin, pentoxifylline and spironolactone on hepatic
fibrosis and portal hypertension in rats with bile duct ligation. J Hepatol
17. Sugano S, Suzuki T, Nishio M, Makino H, Okajima T. Chronic splanch-
nic hemodynamic effects of low-dose transdermal nitroglycerin vs low-
dose transdermal nitroglycerin plus spironolactone in patients with
cirrhosis. Dig Dis Sci 1997;42:529-535.
18. Mironneau J. Calcium channel antagonist effects of spironolactone, an
aldosterone antagonist. Am J Cardiol 1990;65:7K-8K.
19. Sogni P, Soupison T, Moreau R, Le Moine O, Bacq Y, Hadengue A,
Lebrec D. Hemodynamic effects of acute administration of furosemide in
patients with cirrhosis receiving ?-adrenergic antagonist. J Hepatol 1994;
20. Genecin P, Polio J, Groszmann R. Na restriction blunts expansion of
plasma volume and ameliorates hyperdynamic circulation in portal hyper-
tension. Am J Physiol 1990;259:G498-G503.
21. Conn HO, Lindenmuth WW, May CJ, Ramsby GR. Prophylactic porto-
caval anastomosis. A tale of two studies. Medicine 1972;51:27-40.
Propranolol prevents first gastrointestinal bleeding in non-ascitic cirrhotic
patients. J Hepatol 1989;9:75-83.
23. Vorobioff J, Picabea E, Villavicencio R, Puccini V, Rossi O, Bordato J,
Audano M. Acute and chronic hemodynamic effects of propranolol in
unselected cirrhotic patients. HEPATOLOGY 1987;7:648-653.
24. Gerbes A, Remien J, Jungst D, Sauerbruch T, Paumgartner G. Evidence
for down-regulation of beta-2-adrenoceptors in cirrhotic patients with se-
vere ascites. Lancet 1986;1:1409-1411.
25. Escorsell A, Ferayorni L, Bosch J, Garcia-Pagan JC, Garcia-Tsao G, Grace
ND, Rodes J, et al. The portal pressure response to beta blockade is greater
in cirrhotic patients without varices than in those with varices. Gastroen-
survival in cirrhosis. Gastroenterology 1997;113:1632-1639.
27. Merkel C, Marin R. Sacerdoti D, Donada C, Cavallarin G, Torboli P,
Amodio P, et al. Long-term results of a clinical trial of nadolol with or
without isosorbide mononitrate for primary prophylaxis of variceal bleed-
ing in cirrhosis. HEPATOLOGY 2000;31:324-329.
28. Gines P. Uncomplicated ascites. International Ascites Club. Consensus
Document. 5th Meeting. Dallas, TX: AASLD, 1999.
HEPATOLOGY, Vol. 37, No. 2, 2003ABECASIS ET AL.365