Daniel M. Keller, PhD

April 26, 2013

Amsterdam, the Netherlands — In a multivariate analysis of compensated cirrhotic patients infected with hepatitis C virus, spleen stiffness and model for end-stage liver disease (MELD) score emerged as independent predictors of decompensation, leading investigators to develop a predictive model.

"In compensated hepatitis C–related cirrhotic patients, the proposed model…could replace hepatic venous pressure gradient for predicting portal-hypertension-related clinical decompensation," senior author Davide Festi, MD, from the University of Bologna in Italy, told delegates here at the International Liver Congress 2013.

Dr. Festi explained that the prognosis of compensated cirrhotic patients is strongly associated with the development of portal hypertension, and the gold standard for evaluating portal hypertension is the hepatic venous pressure gradient. However, it is an invasive technique that should be performed only by highly experienced operators. Potential complications include bleeding and supraventricular arrhythmias.

Noninvasive tests to predict portal hypertension include serum markers, ultrasonography, elastosonography, magnetic resonance elastography, and transient elastography (FibroScan). Elastography measurements correlate well with hepatic venous pressure gradient.

For their study, Dr. Festi and colleagues examined 85 patients with compensated hepatitis C–related cirrhosis and a normal body mass index. Median age was 59 years, just over half of patients had esophageal varices, and 66% were men.

After study participants were screened with transient elastography for liver and spleen stiffness, esophagogastroduodenoscopy, biochemical tests, abdominal ultrasonography, and hepatic venous pressure gradient, the investigators conducted clinical, biochemical, and ultrasound assessments every 6 months. If patients had a low risk for varices at baseline, they underwent esophagogastroduodenoscopy every 12 months. At follow-up, ascites, variceal bleeding, or hepatic encephalopathy were considered evidence of clinical decompensation.

At baseline, median aspartate aminotransferase and alanine aminotransferase values were each 56 U/L, platelet count was 109.5 × 103/L, median MELD score was 9, liver stiffness was 23 kPa, spleen stiffness was 56 kPa, and hepatic venous pressure gradient was 13 mm Hg. Median follow-up was 730 days. Five patients were lost to follow-up at 2 years, leaving 80 evaluable patients.

At 2 years, 26 of the 80 patients (32%) experienced clinical decompensation, and 11 (14%) experienced other events without decompensation (development of hepatocellular carcinoma, enlargement of varices, or spleen enlargement).

Model Predicts Chance of Decompensation at 2 Years

On multivariate analysis, independent predictors of clinical decompensation were spleen stiffness (hazard ratio [HR], 1.09; 95% confidence interval [CI], 1.04 - 1.13; P < .001) and MELD score (HR, 1.43; 95% CI, 1.07 - 1.91; P = .016). The investigators incorporated these 2 elements into an equation that predicted clinical decompensation at 2 years.

With an upper predicted-risk cutoff of 0.78, the model had a positive predictive value of 88% that a clinical decompensation event would occur within 2 years; with a lower cutoff of 0.2, the model had a negative predictive value of 97% that no event would occur within 2 years.

The investigators developed a decision algorithm for predicting decompensation. If spleen stiffness is less than 53.4 kPa, the patient has a 97% chance that no decompensation-defining event will occur within 2 years. If spleen stiffness is above 53.4 kPa, the equation they developed predicts decompensation with 88% accuracy.

After Dr. Festi's presentation, 2 audience members debated whether spleen stiffness results from congestion or fibrosis. In the end, there was no definitive resolution of the issue. However, regardless of the cause, the model stands.

The association between splenomegaly and portal hypertension has been known for years. "What is new is the idea of measuring spleen stiffness. So far what we have been measuring is spleen enlargement, which, along with platelet count, is usually a good sign of portal hypertension," session chair Laurent Castera, MD, from Hôpital Beaujon and the University of Paris in Clichy, France, told Medscape Medical News.

He explained that liver stiffness not only helps in the staging of liver fibrosis and the diagnosis of cirrhosis; it also helps determine prognosis. Recent studies have suggested that spleen stiffness correlates well with portal hypertension, "maybe even better than liver stiffness alone," he said. This model and algorithm "could be clinically useful because, although hepatic venous pressure gradient is a reference method for portal hypertension, it's only available in very few centers. For instance, in France, there are fewer than 5 centers.... So there's really a role for noninvasive methods to better diagnose and stage portal hypertension," Dr. Castera noted.

However, he advises caution when trying to draw conclusions from the study because of the limited number of patients involved and the short follow-up. Furthermore, from a technical standpoint, measuring spleen stiffness has its limitations "because, when using transient elastography, you cannot choose the region of interest. It's blind," he explained. "Also, it was not initially designed to measure spleen stiffness, but liver stiffness."

In addition, a limitation to measuring liver or spleen stiffness is obesity, which affects "around 20% of patients, at least in Europe, and possibly more in the United States," he said.

Dr. Castera cautions that at this point, the concept of using spleen stiffness is based on a single study. To establish relevant cutoffs and to validate the concept will require larger groups of patients and several independent studies.

With the increasing prevalence of compensated cirrhosis, especially from hepatitis C virus infection, but without signs of portal hypertension, such as varices, "what you want is an exam that is able to rule out very confidently the presence of varices," he noted. "The use of noninvasive methods may be useful because you would spare an invasive endoscopy exam."

Dr. Festi and Dr. Castera have disclosed no relevant financial relationships.

International Liver Congress 2013: 48th Annual Meeting of the European Association for the Study of the Liver (EASL). Abstract 22. Presented April 25, 2013.