Comparison of Shunt Outcomes for Nonbacterial Infection Hydrocephalus with Common Hydrocephalus Etiologies: A Retrospective Case-Control Study

Background: Shunting is an established treatment for hydrocephalus, yet reports on shunt outcomes for nonbacterial infection (NBI) hydrocephalus are limited. Furthermore, comparison of mechanisms and rates of failure for shunted NBI hydrocephalus versus more typical etiologies remains undetermined.

Methods: Patients who underwent shunting for hydrocephalus at 2 centers (1995-2020) were included. Indications for shunting were grouped as "typical" (congenital, posthemorrhagic, normal pressure hydrocephalus, malignancy-related, trauma, and idiopathic) and NBI hydrocephalus (coccidioidomycosis, cryptococcosis, and neurocysticercosis). Rates of shunt malfunction were compared.

Results: There were 261 patients shunted for typical hydrocephalus (48.7% male; age = 50.7 ± 21.7) and 93 patients for NBI hydrocephalus (72.0% male; age = 41.8 ± 13.2). For patients with typical hydrocephalus, 29.5% required ≥1 shunt revision, compared with 64.5% with NBI hydrocephalus (P < 1E-5). Of those with malfunction, NBI shunts required more revision operations (median = 3.0; max = 21) than typical shunts (median = 2.0; max = 6; P < 0.05). The censored median time to shunt failure for NBI hydrocephalus was 26.9 months and was not reached for typical etiologies by 180 months. Multivariate analysis showed shunts for NBI hydrocephalus were significantly more likely to fail (hazard ratio = 2.25; 95% confidence interval = 1.58-3.19). A distal pseudocyst was implicated in 30.0% and 2.6% of shunt failures for NBI and typical hydrocephalus, respectively (P < 1E-5). Sixteen (26.7%) NBI shunt failures required revision to lower-resistance systems compared to 6 (7.8%) typical failures (P < 0.05).

Conclusions: Shunts placed for hydrocephalus secondary to nonbacterial infections are complicated by significantly higher rates of malfunction. These patients are prone to develop distal abdominal pseudocysts and often require revision to low-resistance systems.