Cerebrospinal fluid eosinophilia in children with ventricular shunts.

OBJECT Eosinophils have been reported in children with cerebrospinal fluid (CSF) shunts. The goal of this study was to describe the risk factors, relationship to infection, and clinical significance of CSF eosinophilia in a large group of shunt-treated patients. METHODS The authors performed a retrospective review of data obtained in all patients who underwent ventricular shunt placement or revision at the James Whitcomb Riley Hospital for Children between 2000 and 2004. RESULTS Eosinophils were identified during a follow-up shunt evaluation in 93 (31%) of 300 patients after initial shunt placement. Eosinophilia was statistically related to CSF extravasation (p < 0.0001), shunt infection (p = 0.031), blood in CSF (p < 0.0001), younger age at shunt insertion (p = 0.030), and the diagnosis of posthemorrhagic hydrocephalus (p < 0.0001). Patients with CSF eosinophilia had a higher risk of subsequent shunt failure (p < 0.0001). Analysis was performed using data obtained in a cohort of patients with a total of 130 shunt infections. Cerebrospinal fluid eosinophils were identified in 118 infections (90.8%). The leukocytic and eosinophilic reactions were dependent on the infecting organism. Propionibacterium acnes had a statistically lower CSF leukocyte count but higher differential percentage of eosinophils than the other common pathogens. CONCLUSIONS Cerebrospinal fluid eosinophilia is a relatively common finding in children with shunts. Patients with CSF eosinophilia had an increased risk of shunt malfunction in the present series. Eosinophilia is associated with infection, CSF extravasation, and blood in the CSF. Patients with P. acnes-induced shunt infections have higher eosinophil percentages than are found in infections associated with other common organisms. Therefore, in patients with eosinophilia, extended anaerobic culture studies should be performed with particular attention paid to searching for this pathogen.