Assessing liver fibrosis in patients with transfusion dependent beta thalassaemia - a predictive model

Abstract

Introduction and Objectives Liver fibrosis in β-thalassaemia major is mainly due to transfusion-related iron overload. Transient elastography (TE) is an imaging modality which measures liver stiffness/fibrosis non-invasively. TE is simple, safe and efficient. However, inaccessibility and high-cost hinders its routine use. We designed a predictive model to evaluate liver fibrosis using demographic, anthropometric, biochemical and imaging data. Methods Sixteen patients with transfusion dependent beta thalassaemia were recruited to the study. FBC, LFT, serum ferritin and Transient Elastography (TE) and FerriScan measurements were recorded at the baseline and after two years follow up. Multiple regression model was developed to predict liver fibrosis using demographic, anthropometric, biochemical and imaging data. [age, gender, body mass index (BMI), steatosis score, liver iron content, mean pre-Hb over the last year, no of blood transfusions (lifetime), amount of blood ingested over the last year(ml/kg), amount of elemental iron by transfusions over last year(mg/ kg), serum ferritin, SGOT, SGPT and compliance with iron chelation].Results Of 16, 8 (50%) were females, mean (SD) age, BMI and fibrosis scores were 21(4.3) years, 18.8 (2.8) kgm-2 and 9.7(5.7) kPa respectively. Gender, BMI, SGOT, SGPT, compliance, number of transfusions taken lifetime showed significant association with liver fibrosis. The final model showed a coefficient of determination (R2) of 0.859. According to the model, predicted liver fibrosis is given by;-26.18 - 4.38*male+1.01*BMI - 0.11*SGPT+0.32*SGOT+2.78*compliance (rps)+0.04*no. of transfusions. ConclusionThe suggested model is a reliable tool to predict liver fibrosis in transfusion-dependent β-thalassaemia major patients in resource poor settings.