Cirrhosis and Variceal Bleeding

Today’s morning report case involved a 50-year-old man who presented with melena stools in the context of a significant alcohol history.  He had many stigmata of chronic liver disease, prompting worries of variceal bleeding.  His laboratory values were consistent with decompensated cirrhosis and continued to worsen.

We discussed a lot of valuable learning points:

·      It is important to take a good alcohol history in patients with liver disease.  Be clear with patients and use standardized amounts of alcohol (e.g. 1.5 ounces, one 40 oz bottle of liquor) etc.

·      The diagnosis of cirrhosis in a patient who does not already carry the diagnosis can be challenging.  The gold standard would be a liver biopsy, but we rarely do that for patients for a number of reasons.  Generally, if imaging, laboratory values, and clinical features fit with cirrhosis, and a biopsy wouldn’t change management, we wouldn’t do one.  That said, imaging with ultrasound is only around 90% sensitive, which means that a reasonable proportion of people with this common disease will not have the imaging findings.  Composite indices using Fibroscan technology (pulsed ultrasound wave through the liver to measure stiffness) and laboratory values approach the diagnostic accuracy of imaging.

·      Based on the JAMA rational clinical exam series (Does this patient with liver disease have cirrhosis?) the history including alcohol use, bleeding, fatigue, etc. is not very helpful in the diagnosis of cirrhosis, aside from a previous diagnosis of diabetes (LR + 2.8).  Physical findings can be:

o   Terry nails (LR + 16-22)

o   Gynecomastia (LR + 5.8-35)

o   Distended abdominal veins (LR + 11)

o   Encephalopathy (LR + 10)

o   Decreased body hair (LR + 9)

o   Ascites (LR + 7.2)

o   Facial telangiectasia (LR + 5.9-10)

o   Testicular atrophy (LR + 5.8)

o   Palmar erythema (LR + 5.0)

·      We spoke about some of the laboratory findings in cirrhosis:

o   AST and ALT values are rarely very elevated because the degree of hepatic fibrosis means that the production of those transaminases is quite low

o   Platelet values are one of the first to drop – this is usually due to splenic sequestration from portal hypertension, but can also be due to direct marrow toxicity from alcohol resulting in decreased production

o   The INR and PT also elevate early in liver disease

o   Albumin, produced by the liver is typically low in the course of chronic cirrhosis

o   Cirrhosis impairs the liver’s ability to clear bilirubin to a greater extent than it impairs its ability to conjugate it – this can result in a mixed hyperbilirubinemia

o   Hepatic gluconeogenesis is one of the last functions of the liver to be damaged by cirrhosis – hypoglycemia indicates a significant degree of liver failure

o   Keep in mind that “liver function tests” like the albumin, bilirubin, platelets and INR should be distinguished from transaminases like the ALT and AST which have very little to do with function

o   Also keep in mind that the INR, which was designed to monitor anticoagulation in warfarin, does not reflect the degree of anticoagulation in liver disease.  Because hepatic genesis of protein C and protein S is also affected, patients with liver disease have a balanced coagulopathy wherein their INR of even 2.5 may indicate normal balance of thrombosis and antithrombosis.  Unfortunately, hard cutoffs of INRs and platelet levels are enforced for any invasive procedures due to medico-legal concerns regarding bleeding, even though these patients may be substantial pro-thrombotic.

o   From a laboratory perspective, there are values that have high positive likelihood ratios for cirrhosis in the JAMA rational clinical exam series:

§  Platelet count < 110,000 (LR + 9.8)

§  INR prolonged (LR + 5.0)

§  Albumin < 35 (LR + 4.4)

§  Abnormal bilirubin values and transaminases were less useful

·      We spoke about variceal bleeding.  Do not underestimate the ability of these patients to become unstable.  Always have these patients in a monitored setting with large bore IV access instituted.  Orthostatic vital signs and tachycardia may be subtle indicators of a high degree of blood loss, even prior to a change in the hemoglobin value.

·      The pre-endoscopic treatment of variceal bleeding really has four components that must be accomplished in parallel:

o   Stabilization – ABCs and resuscitation fluids.

o   Acid suppression – this is usually accomplished with parenteral proton pump inhibitors.  The goal immediately is to increase gastric pH to promote platelet aggregation and fibrin deposition as the clotting process is heavily impaired in acid environments.

o   Octreotide – this is a somatostatin agonist and reduces portal pressure directly.  This reduces the severity of variceal bleeding.

o   Antibiotics – Patients with cirrhosis, ascites, and upper GI bleeding have very high probabilities of developing spontaneous bacterial peritonitis which can have mortality implications.  Patients should be treated with ceftriaxone 1g Q24h until discharge – doses should be increased to 2g if SBP is documented on paracentesis.  All patients with decompensated liver disease should have a diagnostic paracentesis performed on admission to hospital.

Further Reading:

McGee, S., Abernethy III, W. B., & Simel, D. L. (1999). Is this patient hypovolemic?. Jama, 281(11), 1022-1029.

Thursz, M. R., Richardson, P., Allison, M., Austin, A., Bowers, M., Day, C. P., ... & Hood, S. (2015). Prednisolone or pentoxifylline for alcoholic hepatitis. New England Journal of Medicine, 372(17), 1619-1628.

Lucey, M. R., Mathurin, P., & Morgan, T. R. (2009). Alcoholic hepatitis. New England Journal of Medicine, 360(26), 2758-2769.

Williams, J. W., & Simel, D. L. (1992). Does This Patient Have Ascites?: How to Divine Fluid in the Abdomen. Jama, 267(19), 2645-2648.

Udell, J. A., Wang, C. S., Tinmouth, J., FitzGerald, J. M., Ayas, N. T., Simel, D. L., ... & Yoshida, E. M. (2012). Does this patient with liver disease have cirrhosis?. Jama, 307(8), 832-842.

Dever, J. B., & Sheikh, M. Y. (2015). Review article: spontaneous bacterial peritonitis–bacteriology, diagnosis, treatment, risk factors and prevention. Alimentary pharmacology & therapeutics, 41(11), 1116-1131.

Simonovský, V. (1999). The diagnosis of cirrhosis by high resolution ultrasound of the liver surface. The British journal of radiology, 72(853), 29-34.