118

in a randomized, controlled trial, found that the use of variceal ligation instead

of sclerotherapy had a lower failure rate (4% vs. 15%; p = 0.02) and a lower

transfusion requirement (p = 0.05). No statistically significant differences were found

in mortality. Adverse effects (e.g., aspiration pneumonia, esophageal bleeding,

ulceration, and chest pain) occurred in 28% of patients receiving sclerotherapy and

14% with ligation (RR, 1.9; 95% CI, 1.1–3.5; p = 0.03).

118

In another similar but

smaller study, Sarin et al.

119

found that the rate of re-bleeding was lower in the EVL

group than in the sclerotherapy group (6.4% vs. 20.8%; p <0.05). EVL is the

recommended form of therapy for acute esophageal variceal bleeding, although

sclerotherapy may be used in the acute setting if ligation is technically difficult.

Endoscopic treatments are best used in combination with pharmacologic therapy,

which preferably should be started before endoscopy.

111 The AASLD Practice

Guidelines Committee and the Practice Parameters Committee of the American

College of Gastroenterology (ACG) on the Prevention and Management of

Gastroesophageal Varices

p. 548

p. 549

and Variceal Hemorrhage in Cirrhosis recommend that for the control and

management of acute hemorrhage, the combination of vasoconstrictive pharmacologic

therapy and variceal ligation is the preferred approach.

92 Therefore, C.V. should

immediately be treated with octreotide for 3 to 5 days, and EVL to control the

esophageal variceal bleeding.

24,92

BALLOON TAMPONADE

Balloon tamponade is used to control bleeding by direct compression to the site of

origin in massive hemorrhage cases (Table 25-2). It is important to remember that

balloon tamponade is only a temporary measure and can cause pressure necrosis

after 48 to 72 hours. Thus, the balloon should be deflated after 12 to 24 hours.

Deflation and removal of the tube can result in removal of the fibrin scab at the

bleeding site, resulting in re-bleeding. Balloon tamponade will achieve temporary

control of the bleeding and allow time for other measures (e.g., EVL or

sclerotherapy) to be undertaken.

92,110,120,121

ALTERNATIVE TREATMENT MODALITIES

Transjugular Intrahepatic Portal Systemic Shunt

Although the combination of pharmacotherapy (octreotide, terlipressin, or

somatostatin) and endoscopic procedures (EVL or sclerotherapy) has been shown to

be beneficial in controlling acute bleeding, re-bleeding episodes can occur.

92,110 Poor

responders to initial therapy may require further intervention to lower portal pressure

and control the bleeding.

Henderson et al.

122 conducted a prospective, multicenter trial comparing distal

splenorenal shunt (DSRS) to TIPS for variceal bleeding refractory to medical

treatment with β-blockers and endoscopic therapy. Patients with Child–Turcotte–

Pugh class A and class B cirrhosis and refractory variceal bleeding were randomly

assigned to DSRS or TIPS. No significant differences were found in survival at 2 and

5 years (DSRS, 81% and 62%; TIPS, 88% and 61%, respectively). Thrombosis,

stenosis, and re-intervention rates were significantly higher in the TIPS group

(DSRS, 11%; TIPS, 82%; p < 0.001). Re-bleeding, encephalopathy, ascites, need for

transplantation, quality of life, and costs did not significantly differ between

groups.

122 TIPS has the advantage of being less invasive and faster than surgical

portal systemic shunts. Long-term patency of the TIPS remains problematic (Table

25-2). TIPS can be used as a bridge to liver transplantation, and might be an effective

option for nonsurgical patients or those with advanced cirrhosis (Child–Turcotte–

Pugh class C) with recurrent bleeding, uncontrolled by pharmacologic and

endoscopic therapy.

86,92,110,123,124 TIPS would be an option for C.V. if EVL and

pharmacologic therapy fail.

Surgery

Surgical creation of a portacaval shunt has been effective in reducing portal pressure

and in preventing recurrent bleeding. These shunts, however, are associated with a

high incidence of hepatic encephalopathy and may exacerbate hepatic parenchymal

dysfunction by shunting blood away from the liver. Mesocaval shunts and distal

splenorenal shunts are also effective in preventing variceal re-bleeding and may be

associated with a lower incidence of hepatic encephalopathy.

125

INFECTION PROPHYLAXIS: SHORT-TERM ANTIBIOTICS

CASE 25-2, QUESTION 4: Should C.V. receive prophylaxis for bacterial infections?

Variceal hemorrhage is a risk factor for the development of severe bacterial

infections, which increase mortality.

92 Short-term administration of antibiotics for the

prevention of bacterial infections in patients with variceal hemorrhage has

demonstrated favorable results.

126,127

In a prospective, randomized trial comparing

norfloxacin 400 mg BID for 7 days (n = 60) with no treatment controls (n = 59), the

norfloxacin group had a significantly lower incidence of spontaneous bacterial

peritonitis (SBP, 3.3% vs. 16.9%; p < 0.05); although the decrease in mortality

(6.6% vs. 11.8%) did not reach statistical significance.

126 Because of the emergence

of infections caused by quinolone-resistant bacteria, Fernandez et al.

127 compared

oral norfloxacin versus IV ceftriaxone in the prophylaxis against bacterial infection

in cirrhotic patients with GI hemorrhage. Patients were randomly assigned to oral

norfloxacin 400 mg BID or IV ceftriaxone 1 g/day for 7 days. Antibiotics were

initiated after emergency endoscopy and within 12 hours of hospital admission. The

probability of developing proven infections (26% vs. 11%; p < 0.03), and

bacteremia or spontaneous bacterial peritonitis (12% vs. 2%, p < 0.03) was

significantly higher in patients receiving norfloxacin as compared with ceftriaxone.

No significant difference was seen between groups in mortality at 10 days.

127 A

Cochrane review of 12 trials evaluating the role of antibiotic prophylaxis with

placebo or no antibiotic prophylaxis to prevent bacterial infections in cirrhotic

patients with upper gastrointestinal bleeding reported that antibiotic prophylaxis was

associated with a significant decrease in mortality from bacterial infections (RR,

0.43; 95% CI, 0.19–0.97).

128

The AASLD guidelines recommend 7 days of antibiotic prophylaxis for prevention

of SBP in patients with variceal hemorrhage with oral norfloxacin (400 mg BID) or

ceftriaxone IV (1 g/day; when oral administration is not possible).

38 C.V. should be

treated with norfloxacin 400 mg orally once daily (dose adjusted for creatinine

clearance of 30 mL/minute), or ceftriaxone 1 g/day for 7 days to prevent SBP.

PRIMARY PROPHYLAXIS

CASE 25-2, QUESTION 5: All variceal hemorrhage interventions up to this point were aimed at terminating

the acute bleeding episode. Could drug therapy have helped prevent the first episode of bleeding from C.V.’s

esophageal varices?

Preventing the initial occurrence of variceal bleeding is referred to as primary

prevention or primary prophylaxis. Pharmacologic prophylaxis is aimed at reducing

the HVPG to ≤12 mm Hg, or a decrease from baseline of ≥20%.

129,130

In a small study

by Vorobioff et al.,

129 none of the patients with HVPG ≤12 mm Hg bled from portal

hypertensive-related causes as compared to 42% in the HVPG >12 mm Hg group. In

addition, only one of the six patients with a HVPG less <12 mm Hg as compared with

16 (of 24) in the HVPG >12 mm Hg group died during the study period (p < 0.06).

129

In addition, Escorsell et al.

130 confirmed that a fall of HVPG by 20% or more from

baseline was associated with a decreased risk of variceal bleeding (6% vs. 45%; p =

0.004).

β-Blockers

Nonselective β-adrenergic blockers decrease portal pressure through a reduction in

portal venous inflow as a result of a decrease in cardiac output (β1

-adrenergic

blockade) and splanchnic blood flow (β2

-adrenergic blockade). Only nonselective βblockers have an adrenergic vasoconstrictive effect on mesenteric arterioles,

resulting in a decrease in portal pressure. Usual starting dosages of propranolol are

10 mg 3 times a day, or nadolol 20 mg daily. Selective β-blockers (e.g., atenolol and

metoprolol) have little effect on mesenteric arterioles and have not been shown to be

effective in primary prophylaxis.

131

Propranolol or nadolol, given in dosages to reduce the resting heart rate to 55 to

60 beats/minute or by 25%, have been shown to prevent or delay the first episode of

variceal bleeding.

15 Nonselective

p. 549

p. 550

β-blockers are considered first-line drug therapy in the prevention of variceal

hemorrhage based on numerous randomized, placebo-controlled trials and metaanalyses.

110,132 For example, Pascal et al.

133 conducted a prospective, randomized,

multicenter, single-blinded trial of propranolol compared with placebo in the

prevention of bleeding in patients with large esophageal varices without previous

bleeding. Patients received either propranolol or placebo, with the endpoints of the

study being bleeding and death. The dosage of propranolol was progressively

increased to decrease the heart rate by 20% to 25%. The cumulative percentages of

patients free of bleeding 2 years after inclusion in the study (74% vs. 39%; p < 0.05)

and cumulative 2-year survival (72% vs. 51%; p < 0.05) were higher in the

propranolol compared to the placebo group.

133

Sarin et al.

134 conducted a prospective, randomized, controlled trial comparing

EVL plus propranolol with EVL alone as primary prophylaxis for prevention of first

variceal bleeding among patients with high-risk varices. The mean duration of

follow-up for both groups was about 12.2 months (±10.7 months). EVL was

performed at 2-week intervals until obliteration of varices. Propranolol was

administered at a dosage sufficient to reduce heart rate to 55 beats/minute or 25%

reduction from baseline, and continued after obliteration of varices. No significant

differences were seen in the rates of bleeding and survival between groups, although

more patients in the EVL alone group had recurrence of varices (p = 0.03).

134

In a prospective, randomized, double-blind, placebo-controlled trial of

propranolol for the primary prevention of variceal hemorrhage, Abraczinskas et al.