an evaluation for splenic vein thrombosis (often
associated with pancreatitis) should therefore be
undertaken in such cases.
DIAGNOSIS, SCREENING, AND SURVEILLANCE
Esophagogastroduodenoscopy (EGD) is the diagnostic test of choice for varices. In addition to its
diagnostic utility, direct therapy of varices with ligation or sclerotherapy can be performed at the time of
EGD as well. Clinical practice guidelines recommend
that esophageal varices be graded as small or large
(diameter > 5 mm), as surveillance and bleeding
prophylaxis recommendations differ depending on
the size.60 When 3 grades are used (small, medium,
and large), medium-sized varices should be treated
as large for management purposes.
Because of the high prevalence of varices in
patients with cirrhosis, the significant morbidity and
mortality associated with hemorrhage, and the efficacy of prophylactic measures to prevent bleeding,
screening for varices is recommended.60 EGD is
the recommended screening modality. Alternative
noninvasive tests such as transient elastography in
combination with standard laboratory values and
capsule endoscopy have been studied, but are not
sufficiently accurate to replace EGD.68,69
The risk of bleeding is related to the severity of
liver disease and the size and features of the vari-
ces; therefore, recommended intervals for surveil-
lance EGD differ depending on patient character-
istics. For patients with compensated cirrhosis but
without varices, EGD should be performed every
2 to 3 years (Table 5).60 For those with a history of
decompensation (ascites or hepatic encephalopa-
thy), EGD should be performed yearly. EGD should
also be repeated at the time of hepatic decom-
pensation for those with previously compensated
disease. Patients with small varices who are not
given prophylaxis should receive an EGD every 2
years, except following decompensation as noted
The prevention of variceal bleeding can be
achieved through medical, endoscopic, and endovascular approaches that either target the underlying portal hypertension or the varices themselves.
Portal pressure-lowering therapies work by reducing
either portal blood flow or resistance. Nonselective
beta-blockers (nadolol and propranolol) cause mesenteric arteriolar vasoconstriction and thus decrease
portal blood flow, which reduces portal pressure.
Their doses can be titrated based on the resting
heart rate or the hepatic venous pressure gradient.
The heart rate is typically used, as measurement
Table 5. Variceal Bleeding Primary Prophylaxis
None Small Large
Hemorrhage;Risk* Low EGD every 2–3 years EGD every 2 years or beta-blocker Beta-blocker preferred; consider band ligation
*Low-risk patients are Child-Pugh A without red wale marks on varices; high-risk patients are Child-Pugh B/C and/or with red wale marks on varices.
EGD = esophagogastroduodenoscopy.