OBJECTIVE
Despite significant advances in the medical field and shunt technology, shunt malfunction remains a nightmare of pediatric neurosurgeons. In this setting, the ability to preoperatively predict the probability of shunt malfunction is quite compelling. The authors have compared the preoperative radiological findings in obstructive hydrocephalus and the subsequent clinical course of the patient to determine any association with overall shunt outcome.
METHODS
This retrospective study included all pediatric patients (age < 18 years) who had undergone ventriculoperitoneal shunt insertion for obstructive hydrocephalus. Linear measurements were taken from pre- and postoperative CT or MRI studies to calculate different indices and ratios including Evans' index, frontal horn index (FHI), occipital horn index (OHI), frontooccipital horn ratio (FOHR), and frontooccipital horn index ratio (FOIR). Other morphological features such as bi- or triventriculomegaly, right-left ventricular symmetry, and periventricular lucency (PVL) were also noted. The primary clinical outcomes that were reviewed included the need for shunt revision, time interval to first shunt revision, frequency of shunt revisions, and revision-free survival.
RESULTS
A total of 121 patients were eligible for the analysis. Nearly half of the patients (47.9%) required shunt revision. The presence of PVL was associated with lower revision rates than those in others (39.4% vs 58.2%, p = 0.03). None of the preoperative radiological indices or ratios showed any correlation with shunt revision. Nearly half of the patients with shunt revision required early revision (< 90 days of primary surgery). The reduction in the FOHR was high in patients who required early shunt revision (20.16% in patients with early shunt revision vs 6.4% in patients with late shunt revision, p = 0.009). Nearly half of the patients (48.3%) requiring shunt revision ultimately needed more than one revision procedure. Greater occipital horn dilation on preoperative images was associated with a lower frequency of shunt revision, as dictated by a high OHI and a low FOIR in patients with a single shunt revision as compared with those in patients who required multiple shunt revisions (p = 0.029 and 0.009, respectively). The mean follow-up was 49.9 months. Age was a significant factor affecting shunt revision–free survival. Patients younger than 6 months of age had significantly less revision-free survival than the patients older than 6 months (median survival of 10.1 vs 94.1 months, p = 0.004).
CONCLUSIONS
Preoperative radiological linear indices and ratios do not predict the likelihood of subsequent shunt malfunction. However, patients who required early shunt revision tended to have greater reductions in ventricular volumes on postoperative images. Therefore a greater reduction in ventricular volume is not actually desirable, and a ventricular volume high enough to reduce intracranial pressure is instead to be aimed at for long-term shunt compliance.