Case of the Month 9 – Answer

There are typical things to consider in an alcoholic patient that is vomiting:

• Pancreatitis, cholelithiasis, Mallory-Weiss tear, Boerhaave’s Syndrome, PUD, alcoholic gastritis, and alcoholic ketoacidosis.

The patient also has evidence of cirrhosis (elevated INR and previous hepatic nodules) and depressed bone marrow production of cell lines. But the curveball in this case is the syncope. He did not endorse lightheadedness when he presented, and his vitals were relatively normal, but abdominal pain and syncope should trigger a few important diagnoses of which many can be grouped under one general suspicion. If he didn’t have a history of alcohol abuse, what diagnoses should you consider? Is he at risk for anything in particular given his history of cirrhosis?

Given the history and findings, this patient is at high risk for spontaneous hemoperitoneum. Abdominal pain and syncope should trigger suspicion for intraabdominal bleeding whether it be ruptured aortic aneurysm, bleeding tumors, solid organ bleeding, varix rupture, coagulopathy and DIC, perforated ulcers, or a slew of other diagnoses. Here is a broad swath of various causes:

This patient is at risk for bleeding given his signs of liver dysfunction and low platelet counts, although these numbers alone do not alone predict high risk for spontaneous bleeding. However, his likely having cirrhosis does place him at a very high risk, and one study (3) found an incidence of 0.5%; those with severely decompensated cirrhosis and high MELD scores portended the highest risk. Decompensated cirrhosis is determined by the presence of ascites, variceal bleed, encephalopathy, or jaundice in a cirrhotic patient. MELD or Child-Pugh scores may also be used to assess mortality risk overall. For example, this patient’s MELD score was 11 placing him at  6% estimated 3-month mortality, and any score ≥ 10 should be referred to a liver transplant center. Although his score is not in the highest interval, it suggests advanced cirrhosis and was likely a contributing factor in his presentation.

Hemoperitoneum in patients with cirrhosis is largely a consequence of intraperitoneal varix rupture either in the hepatic vessels, mesenteric vessels, or the gastro-splenic system. However, given that EtOH abuse may contribute to ulcer formation and hepatocellular carcinoma, a perforated ulcer or rupture of vascular tumors may also be the source. The rate of bleeding of known varices is 12–15% per year with a mortality rate 15–20% for each bleeding episodes. The management should include aggressive repletion of blood volume, stabilization of vital signs, and possible operative correction of the source of bleeding.

Bedside ultrasonography to look for free fluid is a readily available diagnostic test in most EDs. However simple ascitic fluid vs. blood in the abdomen cannot be reliably identified on ultrasound alone even if the peritoneal fluid appears to be complex. Therefore, the gold standard for diagnosis of hemoperitoneum is to perform paracentesis. A hematocrit > 5% of the peritoneal fluid is more specific for intraabdominal bleeding and more likely to lead to operative repair or angioembolization. Below is one recommended diagnostic strategy (4) for identifying and treating the source of hemoperitoneum in cirrhotic patients:

In conclusion, patients with evidence of cirrhosis are at risk for numerous abdominal pathologies one of which is spontaneous hemoperitoneum. Abdominal pain and syncope should raise suspicion, and both rapid resuscitation and diagnosis should be emphasized to reduce mortality in these patients.

References:

1. http://www.hepatitisc.uw.edu/go/evaluation-staging-monitoring/evaluation-prognosis-cirrhosis/core-concept/all
2. Nitin M. Parmar, Dr. Mrinal D. Patel, Dr. Shubham S. Negi, Dr. Chetan M. Savani, Dr. Nirmal L. Desai, Dr. Adarsh P. Patel. Spontaneous hemoperitoneum. Gujarat Medical Journal. July 2015, Vol. 70 No. 2
3. Hepatobiliary Pancreat Dis Int. 2011 Dec;10(6):644-8. Hemoperitoneum in cirrhotic patients without abdominal trauma or tumor.  Ma YJ1, Chen EQ, Lu JJ, Tan MZ, Tang H.
4. http://www.hindawi.com/journals/hpb/2009/240780/