Patients, history taking and clinical assessment
We retrospectively reviewed patients with possible iNPH who were admitted to the neurological or neurosurgery department ward of Peking Union Medical College Hospital for CSF TT. All patients were diagnosed according to the guidelines for the clinical diagnosis of idiopathic NPH (iNPH) published in 2005 [23] and the Chinese consensus for iNPH [24].
In the CSF TT and iNPH registry protocol, age, sex, past histories, personal histories, and initial and full-blown symptoms for all patients were recorded. Patients also completed the Mini-Mental State Examination (MMSE) [25], Montreal Cognitive Assessment (MOCA) [26], and activities of daily living (ADL) questionnaire [27]. They also underwent a brief neuropsychological battery assessment including the Symbol-Digit Modalities Test [28], Trail Making Test A [29] and the Stroop Color Word Task-C [30]. The iNPH Grading Scale (iNPHGS) [31] was used to rate the severity of each fundamental symptom of iNPH (gait disturbance, cognitive impairment, and urinary incontinence) on a four-point scale after a detailed interview with the patients and caregivers. All subjects underwent a CSF TT and brain MRI.
The CSF TT was performed by lumbar puncture followed by the measurement of CSF opening pressure and drainage of 30 ml of CSF. Before and after the procedure, cognitive function and walking ability were evaluated by the means of the 5-meter up and go test (TUG), the 10-meter walking test, the GPT(Model 32025, Lafayette instrument, USA)and a brief executive function battery, which was previously described [32]. The time of TUG, the time and steps that the patient took to complete the 10-meter walking test and a video of the walking tests were recorded. Additional evaluations were conducted at 8 hours, 24 hours and 72 hours after CSF TT. The following criteria were used to identify responders: 1. CSF TT responders were defined as patients with reductions in their time and number of steps in the walking test at least once after the CSF TT, with decreases of 10% for both parameters or a decrease of 20% on at least one of these parameters [33]. 2. Gait was improved based on videos that were evaluated by two neurologists who were blinded to the patient’s clinical and neuroimaging characteristics. The patients who met 1 of 2 criteria were identified as CSF TT responders.
After the detailed clinical assessment, the patients with neuroimaging support or CSF TT response were suggested to have shunting.
Inclusion and exclusion criteria
The patients were recruited in the study according to the following criteria: 1. the patients completed the multi-time point assessment of walking and neuropsychological tests; 2. the patients had undergone at least 3 GPT sessions, including a baseline session. The exclusion criteria were as follows: 1. patients who were unable to finish GPT for any reason; 2. patients who could not tolerate 30 ml of CSF drainage during the CSF TT.
Brain MRI
Axial and sagittal spin-echo T1-weighted images, axial fast spin-echo T2-weighted, fluid-attenuated inversion recovery images were obtained by using a 1.5-T MRI unit (Signa Excite, General Electric, Milwaukee, WI, USA). Diffusion tensor imaging (DTI) was one selected test for patients. We obtained and analyzed 18 patients’ DTI data. The regions of interest (ROIs) were the bilateral anterior and posterior periventricular white matter that were set as circular areas with a diameter of 2 mm perpendicular to the longitudinal axis of the ipsilateral ventricle on the brain imaging slice with the lateral ventricle (detailed in Fig. 1). The fractional anisotropy (FA) and mean diffusivity (MD) values were measured across ROIs on DTI images. The ROIs were chosen according to our pilot study data [34], which indicated that the FA and MD values in the bilateral anterior and posterior periventricular white matter were correlated with walking ability and cognition in possible iNPH patients. A consistency test was performed (Cronbach's a coefficient > 0.6).
Statistical methods
The continuous variables were described as mean ± standard deviation or median and interquartile range, as appropriate. The scores for various evaluation indexes before and 8 hours, 24 hours and 72 hours after CSF TT conformed to skewed distributions. The differences in the GPT performance among the different time points before and after the CSF TT (i.e., baseline, 8 hours, 24 hours and 72 hours) were analyzed using a nonparametric paired sign rank-sum test. The Bonferroni correction was applied for the repeated measurements, and the statistical significance level was set at p < 0.05/4 = 0.01.
We compared the maximum improvement ratios for the GPT performance between the CSF TT responder and nonresponder groups by means of the Mann-Whitney U test. The improvement ratio was calculated as follows: (baseline GPT performance – GPT performance after CSF TT)/baseline GPT performance. The maximum improvement ratio for GPT was extracted for the different times (i.e., 8 hours, 24 hours, and 72 hours) after the CSF TT.
The GPT and symbol-digit modalities test score were combined into the complex visual motor speed index. The maximum improvement ratios for the complex visual motor speed index after CSF TT were also compared between the responder and nonresponder groups using the Mann-Whitney U test. For 8 hours, 24 hours, and 72 hours after CSF TT, we also correlated the improvement ratio of the GPT performance on each time point with that of the corresponding walking test and the improvement ratio of the complex visual motor speed index on each time point with that of the walking test by means of Spearman’s correlation. According to the literature reported[22], The formula was the improvement ratio of the complex visual motor speed index = (the improvement ratio of symbol-digit modalities test raw score + the improvement ratio of mean bilateral GPT performance)/2. The statistical significance level was set at p < 0.05.
The correlations between the baseline GPT results and the baseline clinical assessments, including iNPHGS scores, walking test results, and neuropsychological performance, were tested by means of Spearman’s correlations. The correlations between the baseline GPT results and the DTI parameters from the periventricular white matter lesions were also analyzed using Spearman’s correlation method. The statistical significance level was set at p < 0.05.
All statistical analyses were performed with the statistical software package SPSS for Windows (version 13.0.; SPSS Inc., Chicago, IL, USA).