Selection of patients with idiopathic normal-pressure hydrocephalus for shunt placement: a single-institution experience

Object. The ability to predict Outcome after shunt placement in patients with idiopathic normal-pressure hydrocephalus (NPH) represents a challenge. To date, no single diagnostic tool or combination of tools has proved capable of reliably predicting whether the condition of a patient with suspected NPH will improve after a shunting procedure. In this paper, the authors report their experience with 120 patients with the goal of identifying CSF hydrodynamics criteria capable of selecting patients with idiopathic NPH. Specifically, they focused on the comparison between CSF-outflow resistance (R-out) and intracranial elastance (IE). Methods. Between January 1977 and December 2005,120 patients in whom idiopathic NPH had been diagnosed (on the basis of clinical findings and imaging) underwent CSF hydrodynamics evaluation based on an intraventricular infusion test. Ninety-six patients underwent CSF shunt placement: 32 between 1977 and 1989 (Group I) on the basis of purely clinical and radiological criteria; 44 between 1990 and 2002 (Group II) on the basis of the same criteria as Group I and because they had an IE slope > 0.25; and 20 between 2003 and 2005 (Group III) on the basis of the same criteria as Group II but with an IE slope >= 0.30. Outcomes were evaluated by means of both Stein-Langfitt and Larsson scores. Patients' conditions were considered improved when there was a stable decrease (at 6- and 12-month follow-up) of at least 1 point in the Stein-Langfitt score and 2 points in the Larsson score. Results. Group I: while no statistically significant difference in mean R-out value between improved and unimproved cases was observed, a clear-cut IE slope value of 0.25 differentiated very sharply between unimproved and improved cases. Group II: R-out values in the 2 unimproved cases were 20 and 47 mm Hg/ml/min, respectively. The mean IE slope in the improved cases was 0.56 (range 0.30-1.4), while the IE slopes in the 2 unimproved cases were 0.26 and 0.27. Group III: the mean IE slope was 0.51 (range 0.31-0.7). The conditions of all patients improved after shunting. A significant reduction of the Evans ratio was observed in 34 (40.5%) of the 84 improved cases and in none of the unimproved cases. Conclusions. Our strategy based on the analysis of CSF pulse pressure parameters seems to have a great accuracy in predicting surgical outcome in clinical practice.