Clearing up confusion about cirrhosis

Tips to diagnose and deal with liver disease.


Cirrhosis can often be a point of confusion for hospitalists and trainees, said gastroenterologist Jeffrey S. Crippin, MD, FACP, offering ascites as an example.

“They frequently think of ascitic fluid as pleural fluid. I'll get a call: ‘Dr. Crippin, I checked the ascites and it's a transudate,’” he said. “OK, so which direction did the needle go when you stuck it in the peritoneal cavity?”

Dr. Crippin, a professor and associate chairman for clinical programs in the department of medicine at Washington University School of Medicine in St. Louis, brought laughter to his Hospital Medicine 2017 audience but also useful tips on cirrhosis and liver transplantation.

Jeffrey S Crippin MD FACPslashPhoto courtesy of Dr Crippin
Jeffrey S. Crippin, MD, FACP/Photo courtesy of Dr. Crippin.

He reminded attendees that there are many causes of cirrhosis besides alcohol dependence. “What I frequently get, though, is doctors—not just hospitalists, all specialties—they see ascites [and say] ‘Oh, you've got cirrhosis.’ And then what follows after that? ‘Well, you need to stop drinking,’” Dr. Crippin said.

No matter the cause of cirrhosis, ascites may be present and warrants proper workup, said Dr. Crippin. In patients with a history of cirrhosis and ascites, always perform a cell count with differential and an albumin on ascitic fluid to determine whether it's infected or caused by portal hypertension, respectively, he said.

Determine this by comparing the albumin level in the ascitic fluid to the albumin level in the serum, generating the serum to ascites albumin gradient, or SAAG, a term Dr. Crippin does not like because its imprecision can cause confusion. “To me, it should be a SAAD: a serum to ascites albumin difference,” he said. “I had a trainee years ago … [who] had given me a ratio rather than a difference.”

If the difference is equal to or greater than 1.1 g/dL, portal hypertension is the cause of ascites (its most common etiology), whereas a lower figure indicates another cause, Dr. Crippin said. In about 85% of cases, the portal hypertensive patient will have cirrhosis, but don't forget about patients with severe cardiac dysfunction, he warned.

“I've had patients referred to me for liver transplant evaluation for refractory ascites, and then you look at the history, and there's some congenital heart defect with severe right ventricular dysfunction that is causing ascites,” Dr. Crippin said. “Although there's what's often referred to as cardiac cirrhosis, the liver was an innocent bystander.”

He outlined the small group of diseases associated with a low serum to ascites albumin difference, including atypical intra-abdominal infection (e.g., tuberculous peritonitis) and peritoneal malignancy, which often occur with such clinical symptoms as fever and wasting.

However, it is common to “get sucked into the trap” of relying on a fluid sample to be positive for tuberculosis (TB) or a cytology sample to show that the patient has disseminated cancer, Dr. Crippin said. “The yield rates on TB cultures are relatively low, plus it takes a long time—weeks—to get those back,” he noted. “And cytology, even in the presence of peritoneal carcinomatosis, is only going to be positive in around 50% of patients.”

Dr. Crippin polled the audience members, and a majority said they would routinely send a patient to radiology rather than routinely perform a diagnostic paracentesis. “This is a thing that frequently leads to a great deal of procedure-induced irritable bowel syndrome,” he joked. “A physician is nervous about sticking a needle into a body cavity and is worried about horrible, horrible complications. … You can do a paracentesis.”

To perform a diagnostic paracentesis, use a 20- to 22-gauge needle to obtain 30 to 60 mL of fluid for cell count with differential, albumin, and total protein. Obtaining a culture, cytology, or levels of glucose, carcinoembryonic antigen, or amylase is not always necessary, he said. “You can always tap them again if that gives you additional information,” Dr. Crippin explained, adding that risk of bleeding is minimal (about 1 out of 1,000).

For a culture sent to the microlaboratory, the yield is about 50% for a positive result when there are known bacteria in the fluid, Dr. Crippin noted. “If you do bedside inoculation of blood culture bottles, that yield goes up to 80%,” he said. “Big difference, isn't it? I would encourage you, if you're going to do this test and going to send off fluid for culture, do bedside inoculation.”

Check glucose or carcinoembryonic antigen if the clinical scenario fits a potential perforation, Dr. Crippin said, and check amylase if there is a history of pancreatic ductal injury or pancreatitis. “But for the most part, you can really trim this down to a very limited amount of tests,” he said.

Time from admission to sample is particularly crucial if the patient has spontaneous bacterial peritonitis, Dr. Crippin said. “If you don't make that diagnosis for six or eight or 12 hours, you're missing important time with antibacterials,” he said.

The threshold for a diagnosis of spontaneous bacterial peritonitis is a polymorphonuclear leukocyte count greater than 250 cells/mm3, and antibiotics should be started immediately if this is met, Dr. Crippin said. “They may not have peritonitis, but studies have shown you can save somebody's life by doing this,” he said.

Gram-negative aerobes are the most common offender, and Dr. Crippin recommended treatment with a broad-spectrum antibiotic, such as ceftriaxone or cefotaxime, over the course of five to seven days. He added that studies have shown giving a patient a prophylactic antibiotic at discharge can help prevent additional episodes of peritonitis.

The studies were done with norfloxacin, but that may not be the most practical broad-spectrum antibiotic for patients to obtain, and ciprofloxacin may be 100 times cheaper, Dr. Crippin said. “I still have partners who send patients home on norfloxacin,” he said. “It's usually followed by a phone call from the patient two days later: ‘Well, I went to Walgreens, and they said the norfloxacin was going to be $400. I can't afford that.’”

One episode of spontaneous bacterial peritonitis is associated with a 50% chance of mortality within one to two years. “That patient, if they're a candidate, should be referred to the local transplant center to see if they can actually look at a transplant,” he said.

There are “many, many misunderstandings about liver transplantation,” Dr. Crippin said. Many times, even if the patient does not seem to be a transplant candidate, calling the local transplant center is still worthwhile, he said. “At our place, what I've come to realize is that we oftentimes are going to have more expertise in managing this, so we'll frequently take these patients in a transfer even if transplantation is not in the future,” said Dr. Crippin.

Despite popular belief, alcohol use disorder is not an absolute contraindication to transplantation, he said, adding that most centers will ask the patient to be abstinent for six months in order to show commitment and allow alcoholic hepatitis to improve. “I've gotten calls from hospitalists: ‘I don't know if they're going to live another week or two.’ They stop drinking, and they're alive five years later,” said Dr. Crippin, noting that there is also no absolute age cutoff, and centers may consider a patient in his or her 70s.

Finally, no matter where you practice, you may have liver transplant recipients admitted to your hospital, he said. “So don't be surprised, because they may be hundreds of miles from the transplant center,” said Dr. Crippin.

If you are caring for a transplant recipient, relax, take a history and physical, and be a good doctor, he said. “Don't worry as much about whether or not they've got a disease you've never heard of or they're going into rejection,” Dr. Crippin said. “What I want to hear after you evaluate that patient is: What do you think is going on?”