A Pilot Scoring Model of Hepatocellular Carcinoma Screening in Liver Cirrhosis Based on Clinical Symptom and Daily Laboratory Parameter

Abstract

Ni Luh Putu Yunia Dewi, Kadek Mercu Narapati Pamungkas, Putu Itta Sandi Lesmana Dewi, Ni Nyoman Gita Kharisma Dewi, Dwijo Anargha Sindhughosa and I Ketut Mariadi

Background: Hepatocellular carcinoma is persistent in developing countries. However, it remains difficult to diagnose due to lack of access to radiological and specific tumours marker examination.

Aim: To build a simple scoring model for hepatocellular carcinoma screening in liver cirrhosis patients by using clinical and standard laboratory examinations in the rural areas.

Methods: A cross-sectional, analytical observational study was done to collect data of liver cirrhosis and hepatocellular carcinoma patient. A total of 96 hospitalized patients with liver cirrhosis were included in the study. Multivariate regression analysis was conducted to determine independent factors related to hepatocellular carcinoma. The score of each variable was calculated using the (B/SE)/lowest B/SE formula with strong discrimination power. A scoring model was developed and assessed in terms of sensitivity and specificity. The probability of the total score was calculated using the formula 1/1+exp(-y).

Results: Fourteen (14.6%) patients were hospitalized with hepatocellular carcinoma. The model constructed from the ten potential variables. Three variables were included in the model; high platelet-to-lymphocyte ratio (PLR) ≥150 (p<0.001), high AST-to-ALT ratio (de Ritis ratio) ≥2 (p= 0.020), and presence of hematemesis/ melena (p=0.011). Based on the formula, each variable scored 1. The sensitivity and specificity of the model for diagnosis of HCC in liver cirrhosis at a cut-off point ≥1.5 was 85,7% and 79,3%, respectively (AUC=88,4%). The probability was 0.3%, 4.0%, 36.1%, and 88.4%, following the total score of 0, 1, 2, and 3, respectively.

Conclusion: The existence of hepatocellular carcinoma in liver cirrhosis patients can be detected based on this pilot scoring system for rural areas.

PDF