Avaliação da dosagem sérica do ácido hialurônico e do colágeno IV como marcadores de fibrose hepática em pacientes de área endêmica de esquistossomose mansônica no estado da Bahia

Abstract

Noninvasive markers have been used to predict the presence and degree of liver fibrosis in chronic viral hepatitis, cirrhosis and schistosomiasis. This study was designed to compare serumlevels of hyaluronic acid (HA) and collagen IV (C-IV) to ultrasonographic diagnosis of liver fibrosis. 411 subjects from a highly endemic area for schistosomiasis in the Bahia state, in Brazil, have been submitted to clinical and ultrasound (US) examinations. Subjects younger than 18 andthose with an ultrasound diagnosis of cirrhosis were not eligible. Seventy nine subjects were selected for evaluation of fibrosis markers. S. mansoni ova were sought in the stools using the Kato-Katz technique. Serum samples obtained from all participants were stored at -20ºC. Allpatients were tested for viral hepatitis B and C. HA and C-IV serum levels were measured by an ELISA technique (Echelon Biosciences Inc., Salt Lake City, USA). US examination was performed in all patients using the GE Logic 100 (GE Healthcare, Chalfont St. Giles, UK), equipment with a 3.5 MHz convex transducer. Definition of presence and grade of fibrosis were based on WHO patterns for US and on the examiners subjective evaluation of the fibrosis. Patients were separated according to WHO patterns in four groups: no fibrosis (A), light fibrosis (C, D, Dc), moderate fibrosis (E, Ec) or intense fibrosis (F). Fibrosis was defined by subjectiveevaluation as absent, light, moderate or intense. Comparison of serum levels between groups of fibrosis were conducted through the analysis of variances (ANOVA). Test accuracy was evaluated through the Receiver Operator Characteristic (ROC) curve analysis. Independent predictors of HA concentration were determined by multiple regression modeling. 6 patients had intense liver fibrosis on ultrasound examination, 21 had moderate and 23 had light fibrosis; 29 individuals did not have fibrosis. HA levels reached statistically significant difference between subjects with and without fibrosis (p=0.000), and between those with light and intense fibrosis (p=0.029). Highest levels were detected in the intense fibrosis group. The area under the ROC curve (AUC) was 0.89 (0.82 -0.97) for liver fibrosis. Values above 115.4 ng/ml disclosed patients with fibrosis [sensitivity (Se): 98%; specificity (Sp): 64%]. Independent predictors of HA concentration were intensity of fibrosis on subjective evaluation, age and collateral circulation. There was no difference in C-IV values between study groups. Serum HA was useful in diagnosing patients with liver fibrosis in endemic areas for schistosomiasis mansoni with high diagnostic accuracy.