RESUMO
BACKGROUND: Accurate diagnosis of inflammatory bowel disease, in particular the differentiation between ulcerative colitis and Crohn's disease, is important for treatment and prognosis. Several serological markers have been used as non-invasive diagnostic tools in inflammatory bowel disease patients both to differentiate ulcerative colitis from Crohn's disease and to define patient subgroups. AIM: To evaluate the diagnostic accuracy of three serological tests in differentiating ulcerative colitis from Crohn's disease by single or combined use. METHODS: Sera from 51 patients with clinically well-defined ulcerative colitis and 50 patients with clinically well-defined Crohn's disease were analysed. Detection assays for the presence of perinuclear anti-neutrophil cytoplasmatic antibodies (pANCA), antibodies against (ASCA) and serum agglutinating antibodies to anaerobic coccoid rods were studied. Sensitivity, specificity, predictive values and likelihood ratios of each of these serological tests were determined. RESULTS: In supporting the diagnosis of ulcerative colitis, the sensitivity and specificity of the pANCA test were 63% and 86%, respectively. The ASCA test (immunoglobulin A or immunoglobulin G positive) for diagnosing Crohn's disease had a sensitivity of 72% and a specificity of 82%. The sensitivity of antibodies to anaerobic coccoid rods in diagnosing Crohn's disease was 52%, whereas specificity was 90%. A combination of pANCA-positive and ASCA-negative results in the case of ulcerative colitis showed a sensitivity and specificity of 51% and 94%, respectively. However, for ASCA-positive and pANCA-negative results in the case of Crohn's disease, sensitivity was 64% and specificity was 94%. The combination of all three tests increased positive predictive value and specificity to 100% for both ulcerative colitis and Crohn's disease. In Crohn's disease patients, positive pANCA was correlated with colonic involvement. No correlation was found between the presence of any of these antibodies and disease activity, duration and behaviour or medical treatment. CONCLUSIONS: The value of these serological tests in differentiating ulcerative colitis from Crohn's disease is limited when used separately but, by combining two or more tests, the positive predictive value and specificity can be improved substantially. These tests might be of help in studying disease heterogeneity, and may contribute to defining various subgroups of patients with different pathogeneses.
