Use of Galactomannan Antigen and Aspergillus DNA Real-time Polymerase Chain Reaction as Routine Methods for Invasive Aspergillosis in Immunosuppressed Children in Greece

Abstract

PurposeInvasive aspergillosis (IA) remains a critical issue in immunosuppressed patients. Detection of galactomannan antigen (GM) in serum samples is included as a criterion of IA by the European Organization for the Research and Treatment of Cancer/Mycoses Study Group. Nevertheless, Aspergillus DNA detection by polymerase chain reaction (PCR) has not yet been included because clinical data validation is lacking. The present study describes the simultaneous performance of GM and PCR tests as routine methods for IA diagnosis. MethodsDuring the period January 2012 to December 2017, a total of 156 white children hospitalized in a tertiary children’s hospital of Athens (97 boys and 59 girls; age range, 5 months–14 years) were examined as possible cases of IA. Patients were classified into 4 groups based on their underlying diseases: hematologic malignancies (107 of 156 [68.6%]), solid tumors (16 of 156 [10.2%]), primary immunodeficiency (12 of 156 [7.7%]), and hereditary immunodeficiency (21 of 156 [13.5%]). GM detection was made with the Platelia Aspergillus Ag kit (Bio-Rad Laboratories, Hercules, California). Sera with a cut-off index ≥0.5 on at least 2 separate blood collections were considered positive. Serum detection of Aspergillus DNA was conducted with real-time PCR MycAssay Aspergillus assay (Myconostica Ltd, Cambridge, United Kingdom). PCR positivity was determined by using a threshold of 38 cycles in at least 1 serum sample. Four or more successive samples per patient were tested. FindingsOverall, 28 of 156 patients (53 of 744 serum samples) were found positive. Eleven patients were positive using both methods (24 samples). Four children were positive only by PCR (6 samples), whereas 13 (23 samples) were positive only with GM in consecutive samples. Agreement of both methods, GM(+)/PCR(+) or GM(–)/PCR(–), was found in 139 patients (90% of total patients) and 715 samples (96.1% of total samples). The agreement of both methods was found: (1) 85% in patients with hematologic malignancies; (2) 100% in patients with solid tumors; (3) 97.5% in patients with primary immunodeficiency; and (4) 98.8% in patients with hereditary immunodeficiency. Overall disagreement was observed in 17 patients, in which the positive result in any of the 2 methods was estimated as true positive in conjunction with radiologic and other clinical findings. ImplicationsThe combination of GM and PCR, provided high diagnostic accuracy in consecutive samples (twice a week). Clinical, radiologic, and other laboratory findings should be taken into consideration in the evaluation of GM and PCR.