RESUMO
<p><b>OBJECTIVE</b>To screen biomarkers which can be used as an auxiliary method in the diagnosis of Henoch-Schönlein purpura (HSP) and to evaluate their diagnostic values by receiver operating characteristic (ROC) curve analysis.</p><p><b>METHODS</b>A total of 127 children diagnosed with HSP between April 2012 and March 2014 were included in the HSP group and an equal number of healthy children were included in the control group. Twelve parameters, i.e., serum amyloid protein A (SAA), interleukin-6 (IL-6), immunoglobulins (IgA, IgG, IgM, and IgE), C-reactive protein (CRP), white blood cell (WBC) count, complements C3 and C4, anti-streptolysin O, and ferritin, were analyzed. The values of the screened biomarkers for diagnosis of HSP were assessed by ROC curve analysis.</p><p><b>RESULTS</b>The HSP group had significantly higher levels of SAA, IL-6, CRP, WBC, IgA, and IgM than the control group (P<0.05). The areas under the ROC curve of SAA, IL-6, WBC, IgA, and IgM for the diagnosis of HSP were higher than 0.7 (P<0.05). The optimal cut-off values of SAA, IgA, IgM, WBC, and IL-6 for the diagnosis of HSP were 3.035 μg/mL, 1579.5 mg/L, 922.5 mg/L, 8.850 × 10⁹/L, and 7.035 pg/mL, respectively; the corresponding sensitivities of the optimal cut-off values for the diagnosis of HSP were 95.1%, 75.6%, 72.3%, 78.0%, and 63.4%, respectively, and the corresponding specificities were 90.2%, 85.4%, 82.4%, 70.7%, and 80.5%, respectively.</p><p><b>CONCLUSIONS</b>SAA, IgA, IgM, WBC, and IL-6 are valuable biomarkers for clinical diagnosis of HSP and among them SAA seems to be the best one.</p>
