Man With Recently Detected Esophageal Varices

April 2, 2006

After a routine examination reveals heme-positive stool, a 56-year-old manundergoes upper and lower endoscopy. Esophageal varices are found, as wellas internal hemorrhoids and a few gastric erosions.

After a routine examination reveals heme-positive stool, a 56-year-old manundergoes upper and lower endoscopy. Esophageal varices are found, as wellas internal hemorrhoids and a few gastric erosions.

The patient has biopsy-proven cirrhosis that resulted from heavy, prolongedalcohol consumption; however, he no longer drinks. He has never hadencephalopathy, upper GI tract bleeding, or jaundice. His weight is stable, andhis abdominal girth has not increased. His appetite is good. He has no historyof significant heart disease, hypertension, chronic obstructive pulmonarydisease, or diabetes. He takes ibuprofen, as needed, for degenerative diseasein a number of his large and small joints.

The patient's heart rate is 72 beats per minute and blood pressure is110/75 mm Hg; he has no fever. He is alert, lungs are clear, and heart is normal.He has no scleral or mucosal jaundice. The abdomen is not distended,and no fluid wave can be detected. There is no palpable hepatosplenomegalyor abdominal mass. Two spider angiomas are present on the left shoulder.The remainder of the examination is normal.

Results of a hemogram and routine chemistry panel are normal. Bilirubinlevel is 1.8 mg/dL; albumin level is 3.4 g/dL; transaminase levels are normal.Prothrombin time is 14.5 seconds (control, 12 seconds).

Which of the following is the most appropriate managementstrategy at this point?A. Refer him to an interventional radiologist for a transjugular intrahepaticportosystemic shunt (TIPS) procedure.
B. Refer him to a surgeon for placement of a portosystemic shunt.
C. Order a series of endoscopic sclerotherapy sessions to obliteratethe varices.
D. Prescribe ß-blockers and isosorbide mononitrate.
E. Prescribe octreotide.

CORRECT ANSWER: DThe key issue in this case is how best to prevent an initialbleeding episode. Variceal hemorrhage is a commoncomplication of the portal hypertension associated withchronic liver disease; it produces considerable morbidityand can be lethal.

The entity has been extensively studied. Firm datademonstrate that:

  • Variceal hemorrhage develops in 25% to 35% of patientswith cirrhosis.
  • Variceal bleeding accounts for more than 75% of upperGI tract bleeding episodes in such patients.
  • The mortality of a variceal bleeding episode mayapproach 30%.
  • Bleeding recurs in up to 70% of patients.
  • Reported 1-year survival rates in patients with cirrhosiswho experiencevariceal bleedingrange from 32%to 80%.1

The managementapproachis determinedbywhether bleedinghas occurred:

  • Primary preventionstrategiesare used inpatients withvarices that havenot bled.
  • Actively bleeding varices require immediate intervention.
  • Secondary prevention strategies are used in patientswho have survived an acute variceal hemorrhage.1

This patient requires primary prophylaxis.

His esophageal varices were detected when endoscopywas ordered to investigate heme-positive stool. Itis doubtful that his heme positivity resulted from an otherwisesilent variceal bleed. Many authorities recommend"screening" endoscopy for all patients with cirrhosis--regardlessof whether they have heme-positive stool--becauseof the high risk of hemorrhage and the availabilityof effective primary prevention.

The 2 chief prophylactic strategies are pharmacologicand endoscopic. Drug therapy is currently preferred forfirst-line prevention. ß-Blockers are used alone or in combinationwith nitrates to reduce portal pressure and, as aresult, variceal pressure.

Studies have demonstrated that when hepatic veinpressures are lowered to less than 12 mm Hg or reducedby 20% of the baseline measurement, the incidence ofhemorrhage is markedly reduced.2 ß-Blockers and nitrates(choice D) can lower portal pressure to this degreein at least some patients and typically reduce the incidenceof bleeding by about 50%.3 They are the agents ofchoice in patients for whom they are not contraindicated.

Pharmacologic prophylaxis has been shown to be atleast as effective as sclerotherapy (choice C) and shuntsurgery (choice B) and is far more cost-effective.4

Octreotide (choice E) is a synthetic vasopressin thatproduces vasoconstriction in the splanchnic bed, which inturn lowers portal pressure and thus stops active varicealbleeding in 80% of patients.5 Because it can be given inemergent settings--without the need for specialized endoscopicor radiologic facilities or personnel--octreotide iscurrently first-line therapy for active variceal bleeding.However, its short half-life and intravenous administrationroute make it unsuitable for prophylaxis. Moreover, fordefinitive results, administration of octreotide must be followedby an endoscopic procedure--either sclerotherapyor band ligation. This patient is not actively bleeding andthus is not a candidate for octreotide at this time.

In a TIPS procedure (choice A), a stent is placed toconnect a hepatic vein with a large branch of the portalsystem. Successful TIPS procedures result in hemodynamicimprovements comparable to those resulting from surgicalshunt placements. However, as with surgical shunting,worsening hepatic encephalopathy and shunt occlusion arecommon complications. Thus, TIPS procedures are currentlyreserved for patients with recurrent variceal bleedingand those with refractory acute bleeding.

Outcome of this case. A regimen of nadolol and isosorbidemononitrate was prescribed. At 1 year, the patientcontinues to tolerate treatment well and has had no bleedingevents.


REFERENCES:1. Pryor DB, Shaw L, McCants CB, et al. Value of the history and physical inidentifying patients at increased risk for coronary artery disease. Ann InternMed. 1993;118:81-90.
2. Williams SV, Fihn SD, Gibbons RJ. Guidelines for the management of patientswith chronic angina: diagnosis and risk stratification. Ann Intern Med.2001;135:530-547.
3. Mark DB, Shaw L, Harrell FE, et al. Prognostic value of a treadmill exercisescore in outpatient with suspected coronary artery disease. N Engl J Med. 1991;325:849-853.
4. Lee TH, Boucher CA. Noninvasive tests in patients with stable coronaryartery disease. N Engl J Med. 2001;344:1840-1845.