Failure of an intact ventriculoperitoneal shunt, in the absence of an
overt infection, is often due to its occlusion by cellular debris and/
or an abdominal pseudocyst. This failure is thought to be caused by an
infection by an organism which is difficult to culture or by some poo
rly defined allergic response to the shunt materials. Little attention
has been directed to the treatment that the shunts receive prior to i
mplantation: specifically, their exposure to ethylene oxide as a means
of sterilization. We have found ethylene oxide metabolites in the spi
nal fluid of children with shunt malfunction months after their system
s were implanted. Many of these patients had coincident CSF eosinophil
ia. In addition, two of the children had detectable serum IgE antibody
directed against an albumin-ethylene oxide conjugated protein. Both o
f these children had several shunt malfunctions within a short period,
yet neither child could be shown to have a shunt infection despite mu
ltiple cultures. We therefore suggest that in some patients proteins a
ltered by ethylene oxide incite an IgE mediated response which may lea
d to shunt malfunction.