This article is more than 5 years old. Information may no longer be current.
Algorithm predicts HCC in patients with cirrhosis
Click Here to Manage Email Alerts
VIENNA — A new algorithm accurately predicted hepatocellular carcinoma among a cohort of patients with cirrhosis, according to a study presented at the 2015 International Liver Congress.
Simona Bota, MD, PhD, of the Vienna HCC Study Group, Medical University of Vienna, Austria, and colleagues retrospectively collected the data of 940 patients with cirrhosis enrolled at the Medical University of Vienna. Hepatovenous pressure gradient measurements, laboratory parameters and tumor markers were measured among the patients and 680 were included and divided into two groups for the final analysis: training cohort of 379 patients enrolled from 2004-2009 and validation cohort of 301 patients from 2010-2014.
Overall, the HCC incidence rate among the training cohort was 8.7% (median follow-up: 3.75 years) and 6.3% among the validation cohort (median follow-up: 1.75 years).
Univariate analysis showed in the training cohort that the presence of clinically significant portal hypertension (P = .007), platelet count less than 100,000 per mm3 (P < .0001), alpha-fetoprotein (AFP) greater than 10 ng/ml (P = .002), esophageal varices (P = .03) and age 50 years or more (P = .04) were associated with HCC occurrence. According to Bota, the following were not associated with HCC occurrence: gender, liver cirrhosis etiology, presence of diabetes mellitus, HIV coinfection, BMI, Child-Pugh class, MELD score, presence or history of ascites, presence or history of hepatic encephalopathy, aminotransferases values, gamma-glutamyl transferase, alkaline phosphatase, bilirubin and albumin.
Multivariate analysis showed that platelet count less than 100,000 per mm3, age 50 years or more and AFP greater than10ng/ml reached statistical significance. Based on the odds-ratio, platelet count less than 100,000 per mm3 was assigned two points and age greater than 50 years and AFP of more than10 ng/ml were assigned one point. The patients were then classified in three groups: group 1 was 0-1 points, group 2 was 2-3 points and group 3 had 3–4 points. The HCC incidence per year in the three groups in the training cohort was 0.6%, 3.2% and 9.2% and 0.3%, 5.5% and 9.2% in the validation cohort (P < .0001 for both), respectively.
“The proposed algorithm can identify the patients with very low, moderate and high risk of HCC development, which could be used for more targeted screening approaches,” Bota said. – by Melinda Stevens
For More Information:
Bota S. Abstract O050. Presented at: International Liver Congress; April 22-26, 2015; Vienna.
Disclosure: The researchers report no relevant financial disclosures.