ABSTRACT
BACKGROUND
Periportal fibrosis is the major pathological consequence of the Schistosoma mansoni infection.
OBJECTIVE
To evaluate the accuracy of serum markers and to construct an index to assess fibrosis.
METHODS
Patients (n=116) with schistosomiasis were evaluated by ultrasound scan and measurements of serum levels of aminotransferases, γ-glutamyl transferase, alkaline phosphatase, hyaluronic acid, cytokines and platelets. Ultrasound images were used to evaluate the fibrosis using Niamey's classification and identified 19 patients without periportal fibrosis (patterns A and B), 48 with mild to moderate fibrosis (C and D) and 49 with advanced fibrosis (E and F).
RESULTS
Using multivariate analysis, a model was created, which involved alkaline phosphatase and platelets and could separate patients with different patterns of fibrosis. This index showed a better performance in separating patients without fibrosis from with advanced periportal fibrosis. The biological index showed an area under the ROC curve of 1.000. Using values below the lowest or above the highest cut-off point, the presence or absence of advanced fibrosis could be predicted in all patients.
CONCLUSION
The index constructed can be used to separate patients with different patterns of periportal fibrosis, specially to predict advanced fibrosis in schistosomiasis patients.
Headings
Schistosomiasis mansoni; Fibrosis; ROC curve analysis
Thumbnail
