RESUMO
BACKGROUND: Shunt obstruction is a common cause of shunt failure in the treatment of hydrocephalus. Valve occlusion is traditionally believed to originate from elevated CSF protein or cellular components, although detailed evidence is scarce and contradictory. Therefore, this study aimed to examine CSF protein and cell count as risk factors for valve obstruction. METHODS: We retrospectively examined 274 patients who underwent shunt placement for hydrocephalus between 2009 and 2018 and had at least 1 year follow-up. Age, aetiology of hydrocephalus, valve type, occurrence of revision, reason for revision and CSF protein and cell count at the time of shunt insertion and revision surgery were analysed. RESULTS: Thirty-two of 274 patients (11.7%) required revision surgery due to valve occlusion. Mean time to revision was 143 days. CSF white blood cell (WBC) count but not protein was associated with valve occlusion overall. Of all obstructed valve patients, 25% showed CSF protein level within the normal range, whereas 13.6% of the patients overall showed greatly elevated CSF protein level without evidence of valve obstruction. Persistently elevated CSF protein level at the time of shunt revision was significantly associated with valve obstruction within 90 days of initial insertion (early occlusion). Children with congenital malformations and post-haemorrhagic patients were significantly overrepresented in the occlusion group, particularly in the early occlusion group. CONCLUSION: Pathological CSF values such as WBC count and persistently elevated protein level serves as a risk factor for early valve obstruction. Late obstruction occurs independent of normal CSF values. Infants are particularly prone to early and late valve obstructions. CSF protein level at shunt insertion is not predictive of valve occlusion.