Diagnostic performance of galactomannan antigen testing in cerebrospinal fluid

Abstract

Testing cerebrospinal fluid (CSF) for the presence of galactomannan (GM) antigen may help in diagnosing cerebral aspergillosis (CA). However, the use of the CSF GM test as a diagnostic test has been little studied. We evaluated its diagnostic performance by comparing the CSF GM optical density indexes (ODI) at different cutoffs in patients with probable and proven CA to those in patients without CA. Patients from 2 tertiary referral hospitals with suspected CA between 2004 and 2014 and in whom CSF GM ODI had been determined were selected. European Organization for Research and Treatment of Cancer/Invasive Infectious Diseases Study Mycoses Group (EORTC/MSG) definitions of invasive aspergillosis and CA were used, but with the exclusion of the test to be validated (i.e., the CSF GM test) as a microbiological EORTC/MSG criterion. The study population consisted of 44 patients (4 with proven CA, 13 with probable CA, and 27 with no CA). Of the 17 patients with CA, 15 had a CSF GM ODI of>2.0. Of 27 patients without CA, 26 had a CSF GM ODI of<0.5 and 1 had a CSF GM ODI of 8.2. When a GM CSF ODI cutoff of 1.0 was used, the sensitivity, specificity, and positive and negative predictive values were 88.2%, 96.3%, 93.8%, and 92.9%, respectively. The same results were found when a CSF GM ODI cutoff of 0.5 or 2.0 was used. Testing GM in CSF has a high diagnostic performance for diagnosing CA and may be useful to diagnose or virtually rule out the infection without the need for a cerebral biopsy.