Slope until reaching the plateau: a new predictor of valve response obtained by lumbar infusion test for idiopathic normal pressure hydrocephalus.

Abstract

Diagnosis of idiopathic normal pressure hydrocephalus (iNPH) is based on clinical, radiological, and hydrodynamic data of cerebrospinal fluid (CSF) obtained by invasive methods such as lumbar infusion test, which is used to determine the resistance to CSF outflow (Rout). However, Rout has limitations, and its value as predictor of valve response is questioned. Other variables can be obtained by lumbar infusion test, such as the time to reach the plateau (TRP) and the slope until reaching the plateau (SRP). The objectives were to determine if SRP could be a predictor of response to ventriculoperitoneal shunt (VPS) and what variable (Rout versus SRP) would have greater predictive value. Patients with probable iNPH who underwent a lumbar infusion test and were indicated for a VPS were retrospectively studied. Two groups were established, responders and non-responders. Rout, TRP (period between the start of infusion until reaching the plateau measured in seconds) and SRP ((plateau pressure-opening pressure)/TRP) were obtained. For Rout and SRP, the receiver operating curves (ROC) with its areas under the curve (AUC) were calculated. One hundred ten patients were included, being 86 responders (78.20%). Shunt responders had a significantly greater Rout (17.02 (14.45-20.23) versus 13.34 (12.10-16.28) mmHg/ml/min, p = 0.002) and SRP (0.049 (0.043-0.054) versus 0.031 (0.026-0.036) mmHg/sec, p < 0.001) and smaller TRP (641.28 (584.83-697.73) versus 777.65 (654.03-901.27) sec, p = 0.028) than non-responders. The AUC for SRP was greater than the AUC for Rout (0.763 (95 % CI 0.655-0.871, p < 0.001) versus 0.673 (95 % CI 0.595-0.801, p = 0.008), respectively), but the differences were not significant (p = 0.180). SRP could be considered predictor of response to VPS, and its accuracy tends to be better than Rout. So, this variable may be a useful tool to select shunt candidates among patients with probable iNPH.