The Detection of Hepatocellular Carcinoma Using a Prospectively Developed and Validated Model Based on Serological Biomarkers

Abstract

Background: Hepatocellular carcinoma is a common complication of chronic liver disease (CLD), and is conventionally diagnosed by radiological means. We aimed to build a statistical model that could determine the risk of hepatocellular carcinoma in individual patients with CLD using objective measures, particularly serological tumor markers.Methods: A total of 670 patients with either CLD alone or hepatocellular carcinoma were recruited from a single UK center into a case-control study. Sera were collected prospectively and specifically for this study. A logistic regression analysis was used to determine independent factors associated with hepatocellular carcinoma and a model built and assessed in terms of sensitivity, specificity, and proportion of correct diagnoses.Results: The final model involving gender, age, AFP-L3, a fetoprotein (AFP), and des-carboxy-prothrombin ("GALAD") was developed in a "discovery" data set and validated in independent data sets both from the same institution and from an external institution. When optimized for sensitivity and specificity, the model gave values of more than 0.88 irrespective of the disease stage.Conclusions: The presence of hepatocellular carcinoma can be detected in patients with CLD on the basis of a model involving objective clinical and serological factors. It is now necessary to test the model's performance in a prospective manner and in a routine clinical practice setting, to determine if it may replace or, more likely, enhance current radiological approaches.Impact: Our data provide evidence that an entirely objective serum biomarker-based model may facilitate the detection and diagnosis of hepatocellular carcinoma and form the basis for a prospective study comparing this approach with the standard radiological approaches. Cancer Epidemiol Biomarkers Prev; 23(1); 144-53. Ó2013 AACR.

Can we use HCC risk scores to individualize surveillance in chronic hepatitis B infection?

“…Other than AFP, the Lens culinaris agglutinin-reactive glycoform of AFP (AFP-L3), des-γ-carboxyprothrombin, Golgi protein 73 and osteopontin have been evaluated in the screening population [95,96]. On the whole, these tumor markers have modest performance but may pick up a proportion of AFP-negative HCCs.…”

Current biomarkers for hepatocellular carcinoma: Surveillance, diagnosis and prediction of prognosis

“…The incorporation of clinical variables like age and gender into models based on a combination of biomarkers for HCC detection further improve the predictive performance of these models [39] . A model using a combination of age, gender, AFP, AFP-L3 and DCP estimates the probability to suffer from HCC in an individual patient with chronic liver disease with a sensitivity of 86% for HCC in BCLC stage 0 or A and a sensitivity of 94% for later tumor stages [40] . The diagnostic performance of novel circulating biomarkers and scores is summarized in Table 2.…”

Tumor Markers for Hepatocellular Carcinoma: Simple and Significant Predictors of Outcome in Patients with HCC

“…Although tumor marker levels are not included in the diagnostic criteria for HCC or in the screening recommendations in the guidelines of the American Association for the Study of Liver Diseases or the European Association for the Study of the Liver [15,16], they provide valuable supportive information for diagnosing HCC. Furthermore, a recent study found that the combination of these three tumor markers was useful for diagnosing HCC; this combination had very high sensitivity and specificity for diagnosing HCC without the use of imaging studies [17].…”