Siemens Immulite Aspergillus-specific IgG assay for chronic pulmonary aspergillosis diagnosis

Abstract

Chronic pulmonary aspergillosis (CPA) complicates underlying lung disease, including treated tuberculosis. Measurement of Aspergillus-specific immunoglobulin G (IgG) is a key diagnostic step. Cutoffs have been proposed based on receiver operating characteristic (ROC) curve analyses comparing CPA cases to healthy controls, but performance in at-risk populations with underlying lung disease is unclear. We evaluated optimal cutoffs for the Siemens Immulite Aspergillus-specific IgG assay for CPA diagnosis in relation to large groups of healthy and diseased controls with treated pulmonary tuberculosis. Sera from 241 patients with CPA attending the UK National Aspergillosis Centre, 299 Ugandan blood donors (healthy controls), and 398 Ugandans with treated pulmonary tuberculosis (diseased controls) were tested. Radiological screening removed potential CPA cases from diseased controls (234 screened diseased controls). ROC curve analyses were performed and optimal cutoffs identified by Youden J statistic. CPA versus control ROC area under curve (AUC) results were: healthy controls 0.984 (95% confidence interval 0.972-0.997), diseased controls 0.972 (0.959-0.985), screened diseased controls 0.979 (0.967-0.992). Optimal cutoffs were: healthy controls 15\xa0mg/l (94.6% sensitivity, 98% specificity), unscreened diseased controls 15\xa0mg/l (94.6% sensitivity, 94.5% specificity), screened diseased controls 25\xa0mg/l (92.9% sensitivity, 98.7% specificity). Results were similar in healthy and diseased controls. We advocate a cutoff of 20\xa0mg/l as this is the midpoint of the range of optimal cutoffs. Cutoffs calculated in relation to healthy controls for other assays are likely to remain valid for use in a treated tuberculosis population.