Idiopathic normal pressure hydrocephalus (inph) is an

Transcription

1 J Neurosurg 120: , 2014 AANS, 2014 The callosal angle measured on MRI as a predictor of outcome in idiopathic normal-pressure hydrocephalus Clinical article Johan Virhammar, M.D., 1 Katarina Laurell, M.D., Ph.D., 1,2 Kristina Giuliana Cesarini, M.D., Ph.D., 3 and Elna-Marie Larsson, M.D., Ph.D. 4 1 Neurology and 3 Neurosurgery, Department of Neuroscience, and 4 Department of Radiology, Uppsala University, Uppsala; and 2 Neurology, Department of Clinical Neuroscience, Umeå University, Umeå, Sweden Object. Different neuroimaging biomarkers have been studied to find a tool for prediction of response to CSF shunting in idiopathic normal-pressure hydrocephalus (inph). The callosal angle (CA) has been described as useful in discriminating inph from ventricular dilation secondary to atrophy. However, the usefulness of the CA as a prognostic tool for the selection of shunt candidates among patients with inph is unclear. The aim of this study was to compare the CA in shunt responders with that in nonresponders and clarify whether the CA can serve as a predictor of the outcome. Methods. Preoperative MRI brain scans were evaluated in 109 patients who had undergone shunt surgery for inph during Multiplanar reconstruction was performed interactively to obtain a coronal image through the posterior commissure, perpendicular to the anterior-posterior commissure plane. The CA was measured as the angle between the lateral ventricles on the coronal image. The patients were examined clinically before surgery and at 12 months postoperatively. Results. Shunt responders had a significantly smaller mean preoperative CA compared with nonresponders: 59 (95% CI ) versus 68 (95% CI ) (p < 0.05). A CA cutoff value of 63 showed the best prognostic accuracy. Conclusions. The preoperative CA is smaller in patients whose condition improves after shunt surgery and may be a useful tool in the selection of shunt candidates among patients with inph. ( Key Words normal-pressure hydrocephalus NPH magnetic resonance imaging MRI outcome prognosis dementia Idiopathic normal pressure hydrocephalus (inph) is an increasingly recognized condition among the elderly population and is associated with symptoms of gait impairment, cognitive decline, and urinary incontinence. 1 The symptoms can be reduced by shunt insertion, which leads to improvement in 50% 80% of the patients. 13,15,17,22 When selecting candidates for shunt surgery, prognostic information is gathered from clinical examination, neuroimaging, and invasive CSF testing. 19,20 Radiological findings that support the clinical suspicion of inph include dilated ventricles, effaced sulci at the high cerebral Abbreviations used in this paper: AC = anterior commissure; CA = callosal angle; ICC = intraclass correlation coefficient; inph = idiopathic normal-pressure hydrocephalus; MMSE = Mini Mental State Exam; mrs = modified Rankin Scale; PACS = picture archiving and communication system; PC = posterior commissure; ROC = receiver operating characteristic; TUG = timed up and go; WB3m = walking backward 3 m. convexities, and signs of hyperdynamic CSF flow. 3,6,14,21 These radiological signs are considered useful to differentiate inph from other causes of dementia, 3,12 whereas their role in predicting response to shunting remains unclear. 25 One marker of hydrocephalus, the callosal angle (CA, the angle between the lateral ventricles on a coronal image, Fig. 1 right), was found in early studies using pneumoencephalography to be smaller in patients with NPH than in patients with ventricular dilation secondary to parenchymal atrophy. 16,23 Ishii et al. presented a method for measurement of the CA on MRI. 11 They suggested that patients with inph could be distinguished from those with ventricular dilation due to Alzheimer s disease by a CA of less than 90. However, no previously published studies have clarified the usefulness of the CA as an imaging biomarker for the selection of patients with inph for shunt surgery. In this study, we compared the CA in patients with 178 J Neurosurg / Volume 120 / January 2014

2 The callosal angle in inph Fig. 1. Left: A sagittal image is used to identify the AC-PC plane and the posterior commissure. Right: The callosal angle is measured in the coronal plane through the posterior commissure perpendicular to the AC-PC plane. inph responding to shunt surgery with that in nonresponders with the aim of clarifying whether the CA can serve as a predictor of the outcome. Methods The local ethics committee in Uppsala, Sweden, approved the study. Patients The original patient group for this study included 173 patients with inph who had received a shunt at Uppsala University Hospital between January 2006 and December 2010, had been followed up postoperatively at the same hospital, and had given written informed consent to participate in the study. Thirty-six patients were excluded because they lacked a preoperative MRI examination including a sagittal sequence within 18 months before surgery. Of the 137 patients with a valid preoperative MRI examination, 109 were assessed clinically at both baseline and at the end of the follow-up period, 12 months postoperatively. Thus, the final sample consisted of 109 patients, 58 men (53%) and 51 women (47%), with a median age at the time of shunt surgery of 74 years (range years). The mean time (± SD) between MRI and surgery was 32 ± 18 weeks and the mean time to 12-month follow-up was 54 ± 6 weeks after surgery. A total of 28 patients were lost to follow-up. Six patients died before the 12-month follow-up visit: 1 patient died during the 1st month after surgery due to a myocardial infarction, 5 patients died later than 3 months after surgery 2 due to myocardial infarction, 1 due to bacterial pneumonia with septicemia, 1 due to complications related to preexisting pulmonary fibrosis, and 1 of unknown causes. Fifteen patients were lost to follow-up because of shunt complications and 2 patients because of complications related to comorbidity; 1 patient had moved to a different hospital; and 5 patients were lost to follow-up for unknown reasons. Among the remaining 109 patients, 29 (27%) suffered from different types of shunt complications: 1 patient (1%) had an intracerebral hematoma, 10 patients (9%) had subdural hematomas of which none required evacuation, 2 patients (2%) had a suspected shunt infection which was treated by shunt revision, and 16 patients (15%) underwent J Neurosurg / Volume 120 / January 2014 additional surgery because of proximal or distal catheter failure. Five patients suffered from a comorbidity-related complication that had a negative effect on the results at follow-up: 1 stroke with persisting motor impairment, 1 lung resection, 1 radical cystectomy, 1 amputation of the lower limbs, and 1 femur fracture. Comparing patients with or without shunt complications, there were no significant differences in comorbidity, age, radiological parameters, or any clinical parameter at baseline. All patients had ventricular enlargement (mean Evans index 0.38, range ) and gradually evolving symptoms, including a typical gait disturbance with or without cognitive decline or urinary incontinence. No patient had a history of subarachnoid hemorrhage, meningitis, or other known cause of secondary NPH. CSF tap test and infusion tests were used to aid in the selection of shunt candidates. Twenty-two patients (20%) had suffered from a previous stroke, and 7 (6%) had a history of a myocardial infarction. Eight patients (7%) had a psychiatric diagnosis such as schizophrenia or bipolar disorder, and 26 (24%) were being treated with antidepressant drugs. Other drugs being given were: antihypertensive drugs, 66 patients (61%); levodopa, 7 patients (6%); and insulin, 16 patients (15%). The Callosal Angle Measurements of the CA were performed retrospectively, and the diagnostic procedure or selection to surgery was not affected by information about the radiological parameters. All measurements were performed digitally using a clinical picture archiving and communication system (PACS) (Carestream Health). Ishii et al. 11 suggested that the CA should be measured between the lateral ventricles on a coronal MR image through the posterior commissure (PC), perpendicular to the anterior commissure (AC)-PC plane. Multiplanar reconstruction was performed interactively in the PACS for each patient to obtain such a coronal image (Fig. 1). The measurements were performed on 3D T1-weighted images with an approximate voxel size of mm (or slightly greater). In cases in which such thin images had not been acquired, the measurements were made on an interactively reconstructed coronal image (in the plane described above) from a sagittal scan with slice 179

3 J. Virhammar et al. thickness of 3 5 mm or from a coronal scan acquired in a plane not perpendicular to the AC-PC plane (using the capability to re-angle in the PACS). When a reconstruction was made with these thicker slices, the quality of the reconstructed coronal image was reduced. To evaluate the agreement between these 2 methods, 38 patients who had both high-quality 3D scans and scans with thicker sagittal images were assessed in both ways, with the evaluating researcher blinded to the results of the other method. The first author performed all measurements on the MRI scans after training of measuring on other patients together with an experienced neuroradiologist (last author). To calculate intra- and interrater reliability, the 2 investigators independently evaluated the images of 20 randomly selected patients. The first author repeated the evaluation after 6 months, blinded to the results of the first measurements and clinical outcome. A total of 41 (38%) of the preoperative MRI examinations had been performed at Uppsala University Hospital, and 68 (62%) had been performed at referring hospitals. Ten scans (9%) had been performed on a 3-T scanner, 75 (69%) on a 1.5-T scanner, 16 (15%) on a 1-T scanner, and 8 (7%) on a 0.5-T scanner. To establish the degree of ventricular dilation in our patient population, the Evans index (calculated as the ratio between maximum diameter of the frontal horns of the lateral ventricles and maximum inner diameter of the skull) was measured on transverse MR images in all patients. 7 Clinical Evaluation A multidisciplinary inph team, including a neurologist, a neurosurgeon, and specially trained physiotherapists and occupational therapists, collected clinical data prospectively, according to a standardized protocol. The patients were assessed before surgery and at follow-up 12 months postoperatively. Cognitive function was evaluated with the Mini Mental State Exam (MMSE). 8 Improvement of 2 3 levels in MMSE was considered as a possible improvement in cognition and improvement of 4 or more was considered as a definitive improvement. Continence was evaluated with a 6-grade ordinal scale 10 and an improvement of 1 level or more was defined as an improvement in continence. The motor function was tested with 2 ordinal scales, a gait and a balance scale, 10 but also 3 tests in which time and number of steps were noted. The latter included 10 m walking at self-chosen speed (10 m), timed up and go (TUG), and walking backward 3 m (WB3m). 24,26 An improvement of 1 level or more in either the gait or the balance scale or an improvement of at least 20% in time or number of steps in at least 50% of the 3 timed tests performed (10 m, TUG, and WB3m) was considered as an improvement in motor function. A shunt response was defined as an increase of at least 2 points in a symptom scale that included all 3 symptoms of NPH (Table 1). For comparisons, an increase of at least 1 level in modified Rankin Scale (mrss) score was used as an alternative outcome. 4 TABLE 1: Symptom scale* Symptom Points possible improvement in cognition 1 definitive improvement in cognition 2 improvement in continence 1 improvement in motor function 2 * Response to shunting was defined as an improvement of 2 points or more. Statistics The Shapiro-Wilk test of normality was used to determine the distribution of the variables. Intraclass correlation coefficients (ICCs) were calculated for reliability measurement. The Mann-Whitney test and Pearson s chi-square (or Fisher s exact test when appropriate) were used to test for differences in comorbidity and baseline data between patients with or without complications and between shunt responders and nonresponders. To test for differences in the CA between patients with improved and nonimproved condition, an independent-samples t-test was used. In addition, Spearman s or Pearson s correlation coefficients were used for correlation analyses depending on the normality of the variable tested. Receiver operating characteristic (ROC) curves were calculated and 4 different cutoffs for CA were chosen. The first was selected to match a specificity of 90% and the second was calculated using Youden s index (sensitivity + specificity 1), 2 a theoretical model to calculate an optimal cutoff. The third cutoff was chosen to match a sensitivity of 80% and the fourth was set to 90, the cutoff suggested by Ishii. 11 To examine if the CA was an independent predictor of outcome, a multivariate logistic regression model was performed after first testing all variables in Table 2 in a univariate regression model. Shunt response measured with the symptom scale was used as dependent factor and CA, age, sex, and previous stroke were used as predictor variables in the multivariate analysis. The level of significance was set to p < Analyses were performed using SPSS (IBM SPSS Statistics for Macintosh, Version 21.0, IBM Corp.). Results In 109 inph patients, the preoperative mean CA was 61 (95% CI ) (Table 3). The preoperative CA was significantly smaller in patients who responded to shunt surgery than in nonresponders, 59 (95% CI ) versus 68 (95% CI ) (p < 0.05). There was no significant difference in Evans index between shunt responders and nonresponders. At 12-month follow-up, 82 (76%) of 108 patients tested with the symptom scale were defined as shunt responders (improvement 2 points in the symptom scale, Table 1) and 35 (38%) of 92 tested showed improvement in mrs scores. In one patient, outcome was not determined with the symptom scale but only with mrs. Of 100 patients tested, 21 (21%) improved in MMSE with 2 3 levels (possible improvement in cognition) and 28 (28%) were improved in MMSE by at least 4 levels (definitive improvement in cognition). Seventy-four (71%) of 105 patients tested were improved in motor function, and 36 (38%) of 94 tested were improved in continence. Characteristics 180 J Neurosurg / Volume 120 / January 2014

4 The callosal angle in inph TABLE 2: Characteristics of shunt responders and nonresponders in 108 cases* Characteristic Responders (n = 82) Nonresponders (n = 26) p Value CA ( ) <0.05 mean % CI age (yrs) ns median range sex, M/F 43/39 15/11 ns previous stroke 14 (17%) 8 (31%) ns previous MI 5 (6%) 2 (8%) ns antidepressants 19 (23%) 7 (27%) ns other psychiatric disorder 6 (7%) 2 (8%) ns antihypertensive drugs 50 (61%) 16 (62%) ns L-dopa treatment 5 (6%) 2 (8%) ns insulin treatment 10 (12%) 6 (23%) ns * CA = callosal angle; L-dopa = levodopa; MI = myocardial infarction; ns = not significant. of shunt responders and nonresponders are presented in Table 2. The measurements of the CA showed a high intrarater and interrater reliability, with ICCs of 0.97 (95% CI ) and 0.95 (95% CI ), respectively. Likewise, intrarater and interrater reliability for the Evans index was high, with ICCs of 0.98 (95% CI ) and 0.93 (95% CI ), respectively. The agreement between the CA measured on 3D MRI scans compared with measurements on coronal images reconstructed from 3-mm sagittal images was high, with an ICC of 0.93 (95% CI ). There was a significant difference in preoperative CA between patients defined as shunt responders (Table 1) compared with nonresponders at follow-up: 59 (95% CI ) versus 68 (95% CI ) (p < 0.05). The TABLE 3: Radiological and clinical parameters at baseline and follow-up* Parameter Baseline Value J Neurosurg / Volume 120 / January mo Follow-Up Value CA in degrees, mean (95% CI) 61 (58 64) Evans index, mean (range) 0.38 ( ) MMSE score, median (IQR) 24 (21 27) 26 (24 28) mrs score, median (IQR)) 2 (2 3) 2 (1 3) continence scale score, median 3 (2 4) 2.5 (1 3) (IQR) gait scale score, median (IQR) 4 (3 6) 3 (2 5) 10-m walk score in sec, median (IQR) 19 (14 26) 14 (11 20) * IQR = interquartile range. Significant difference between follow-up and baseline (Wilcoxon signed-rank), p < Significant difference between follow-up and baseline (Wilcoxon signed-rank), p < difference was significant also if the mrs was used to separate responders from nonresponders: 57 (95% CI ) versus 64 (95% CI ) (p < 0.05). When patients with complications were excluded, the difference between shunt responders and non responders was more pronounced: 59 (95% CI ) versus 70 (95% CI ) (p < 0.01, n = 76). All variables in Table 2 were tested in a univariate logistic regression with shunt response as dependent factor, and only CA was significant. In a multivariate logistic regression model, a smaller CA was significantly associated with response to shunting (odds ratio 0.97, 95% CI , p < 0.05), after controlling for the variables age, sex, and previous stroke. There was a weak inverse correlation between the CA and Evans index (r = -0.23, p < 0.05). There was a moderate inverse correlation between the CA and improvements in the balance scale (r = -0.35, p < 0.001) and WB3m (r = -0.30, p < 0.01). There were also weak inverse correlations between the CA and improvements in mrs (r = -0.24, p < 0.05), TUG (r = -0.25, p < 0.05), and 10 m (r = -0.22, p < 0.05) at 12-month follow-up. A CA cutoff value of 63 showed the highest Youden s index (0.33, Table 4). If patients with complications were excluded, the prognostic accuracy was slightly better (Youden s index 0.38). See Table 4 for different cutoffs for CA. Discussion In this study of 109 patients with inph who had undergone shunt surgery, the preoperative CA was smaller in shunt responders compared with those who did not improve after shunt surgery. The result was significant both when outcome of shunt surgery was assessed by a symptom scale and when it was assessed by the level of disability (mrs). The mean CA in this study (61, 95% CI ) was in line with the value reported by Ishii and 181

5 J. Virhammar et al. TABLE 4: Sensitivity and specificity for different cutoffs for the callosal angle to predict a shunt responder CA Cutoff ( ) Sensitivity (%) Specificity (%) Youden s Index* for all patients w/ follow-up w/ symptom scale scores (n = 108) for patients w/o complications (n = 76) * Youden s index = sensitivity + specificity 1. colleagues (66 ± 14 ). The repeatability and reliability was good both in the present study (ICC = 0.95) and the study by Ishii et al. (ICC = 0.98). 11 Although a small CA was described early on as a pneumoencephalographic sign of NPH, 16 the prognostic value of the CA for the outcome of shunt surgery has, to our knowledge, not been published previously. A pneumoencephalographic study of the CA included 20 patients with inph. The 9 patients whose status improved with shunting as well as 7 of the 11 patients whose status did not improve had a frontal horn CA smaller than However, the authors did not describe how improvement after shunting was defined or evaluated. Furthermore the CA was measured more anteriorly than in our study and that of Ishii et al., 11 making comparisons difficult. The authors proposed an appealing theory to explain the small CA in hydrocephalus: The expansion of the lateral ventricles causes an elevation of the corpus callosum until it reaches the level of the falx, whereas the lateral portions of the roof of the lateral ventricles are able to continue to rise, thereby decreasing the CA. 16 The falx has a larger cranio-caudal height posteriorly, so when the ventricles expand, the elevation of the posterior portion of the corpus callosum is discontinued before that of the anterior portion. This results in the CA being smaller when measured at the level of the PC, as in our study, than when calculated as the angle between the frontal horns, as in the pneumoencephalographic studies. Consequently, measurements of the CA should be standardized as described above, since the angle varies with the angulation and position of the slice selected for measurement. 11 Also, the software in the PACS should have the function of interactive multiplanar reconstruction. In another pneumoencephalographic study it was shown that patients with NPH had a smaller frontal horn CA (118 ) than patients with atrophy (135 ). 23 However, there was no significant difference between the mean CA in NPH and the mean CA in obstructive hydrocephalus. 23 In the 2005 American guidelines for inph, 20 it was proposed that a CA greater than 40 is supportive of the diagnosis of inph, which is somewhat confusing in the light of previous results. Unfortunately no reference was provided. In the present study, a CA cutoff value of 63 showed the best prognostic accuracy, with sensitivity of 67% and specificity of 65%. Still, the usefulness of the CA to predict shunt responsiveness in patients with inph was somewhat limited. This limitation is shared with other adjunctive tests in inph, in which the sensitivities and specificities rarely exceed what can be accomplished by an interdisciplinary expert team in the selection of shunt candidates. 28 Possibly a combination of different noninvasive prognostic tests can add more certain prognostic information. For example, other promising MRI techniques have been proposed to select shunt candidates, such as aqueductal stroke volume and measurements of cerebral blood flow. 6,27 Although comparing studies is problematic due to different outcome measures, our finding of 76% shunt responders 1 year after surgery is in agreement with recent publications. 15,18 Patient comorbidity (previous stroke in 20%) and antihypertensive drug use (61% of the patients) were considerable and in line with other studies of patients with inph. 9,15 The rate of shunt complications is high in this study because we considered radiologically verified overdrainage as a complication. Although most such patients have no symptoms, 5 the consequence of overdrainage is often an increase of the opening pressure of the shunt valve to a maximum setting to avoid deterioration. This in turn decreases the effect of the shunt and might limit the clinical improvement at follow-up. 29 When we excluded patients with complications, the difference in the CA between shunt responders and nonresponders was even more pronounced. An explanation for why patients with a large CA in this sample were less likely than those with a small CA to respond to shunt surgery could be that the former group suffered from ventricular dilation secondary also to atrophy and not only due to inph. If the narrowing of the CA in inph is an effect of slightly increased intracranial pressure or a result of increased CSF pulsations remains to be investigated. The correlations between a smaller CA and improvements in the clinical tests also support the hypothesis that patients with a narrow CA are more likely to improve after surgery. Figure 2 shows representative preoperative MR images obtained in a patient with a small CA whose condition improved after surgery and a patient with a large CA whose condition did not improve after surgery. Some limitations should be considered. Because inph had been diagnosed in all the patients in this study and surgery was performed for treatment of that condition, the results cannot automatically be generalized to a less selected population. Dropouts due to complications related to the shunt procedure and preexisting comorbidity contributed to the fact that 28 patients in this study could not be included in the analyses at 12-month followup. This is similar to most studies in the field, and the main results were even more robust when patients with complications were excluded. Although data were collected prospectively in a standardized way, the radiological measurements and analyses 182 J Neurosurg / Volume 120 / January 2014

6 The callosal angle in inph Author contributions to the study and manuscript preparation include the following. Conception and design: all authors. Ac quisition of data: all authors. Analysis and interpretation of data: Virham mar, Laurell, Larsson. Drafting the article: Virhammar, Laurell. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Virhammar. Statistical analysis: Virhammar. Fig. 2. Three-dimensional T1-weighted coronal MR images obtained in 2 patients who underwent ventriculoperitoneal shunt insertion because of inph. Left: This patient had affected gait, continence, and enlarged ventricles with an Evans index of 0.37 and a CA of 110. The patient s condition was not improved 12 months after shunting, and deterioration of cognitive function was evident. Right: This patient had urgency incontinence, was wheelchair bound, and had an Evans index of 0.36 and a small CA of 42. The patient demonstrated markedly improved condition 12 months after shunt surgery and was walking unaided. of clinical data for this study were performed retrospectively. This fact limited the possibility of making corrections when specific MRI sequences or some clinical data were missing. On the other hand, the strength was that one person (J.V.) performed the analyses and that his assessment was in good agreement with that of an experienced neuroradiologist (E.M.L.). Thus the reliability and validity of the analyses should be high. In summary, measurement of the CA is uncomplicated and can be performed by any radiologist using precise definitions of the position and angulation of the coronal measurement slice. In this sample of patients who had undergone shunt surgery for inph, the shunt responders had a smaller CA than the nonresponding patients. The results indicate that the CA is useful in the selection of shunt candidates. However, our findings should be confirmed in prospective studies and compared with other radiological markers. Conclusions In this series of patients with inph, the preoperative CA was smaller in patients whose condition improved after shunt surgery than in those whose condition did not improve. The CA is a noninvasive imaging marker that may be a useful tool in the selection of shunt candidates among patients with inph. Acknowledgments We would like to thank Lisa Wernroth for statistical consultation and Jonna Brorsson for aid in the collection of data. We also thank the staff of our NPH team at Uppsala University hospital for their commitment to all the patients. Finally we thank Selanders Stiftelse for their support. Disclosure The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Katarina Laurell recived a grant from the independent foundation Selanders Stiftelse, Uppsala, Sweden to support this study. J Neurosurg / Volume 120 / January 2014 References 1. Adams RD, Fisher CM, Hakim S, Ojemann RG, Sweet WH: Symptomatic occult hydrocephalus with normal cerebrospinal-fluid pressure. A treatable syndrome. N Engl J Med 273: , Akobeng AK: Understanding diagnostic tests 3: receiver operating characteristic curves. Acta Paediatr 96: , Algin O, Hakyemez B, Taskapilioglu O, Ocakoglu G, Bekar A, Parlak M: Morphologic features and flow void phenomenon in normal pressure hydrocephalus and other dementias: are they really significant? Acad Radiol 16: , Banks JL, Marotta CA: Outcomes validity and reliability of the modified Rankin scale: implications for stroke clinical trials: a literature review and synthesis. Stroke 38: , Boon AJ, Tans JT, Delwel EJ, Egeler-Peerdeman SM, Hanlo PW, Wurzer HA, et al: Dutch Normal-Pressure Hydrocephalus Study: randomized comparison of low- and medium-pressure shunts. J Neurosurg 88: , Bradley WG Jr, Scalzo D, Queralt J, Nitz WN, Atkinson DJ, Wong P: Normal-pressure hydrocephalus: evaluation with cerebrospinal fluid flow measurements at MR imaging. Radiology 198: , Evans WA: An encephalographic ratio for estimating ventricular enlargement and cerebral atrophy. Arch Neurol 47: , Folstein MF, Folstein SE, McHugh PR: Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12: , Hashimoto M, Ishikawa M, Mori E, Kuwana N: Diagnosis of idiopathic normal pressure hydrocephalus is supported by MRI-based scheme: a prospective cohort study. Cerebrospinal Fluid Res 7:18, Hellström P, Klinge P, Tans J, Wikkelsø C: A new scale for assessment of severity and outcome in inph. Acta Neurol Scand 126: , Ishii K, Kanda T, Harada A, Miyamoto N, Kawaguchi T, Shimada K, et al: Clinical impact of the callosal angle in the diagnosis of idiopathic normal pressure hydrocephalus. Eur Radiol 18: , Ishii K, Kawaguchi T, Shimada K, Ohkawa S, Miyamoto N, Kanda T, et al: Voxel-based analysis of gray matter and CSF space in idiopathic normal pressure hydrocephalus. Dement Ger iatr Cogn Disord 25: , Kahlon B, Sjunnesson J, Rehncrona S: Long-term outcome in patients with suspected normal pressure hydrocephalus. Neurosurgery 60: , Kitagaki H, Mori E, Ishii K, Yamaji S, Hirono N, Imamura T: CSF spaces in idiopathic normal pressure hydrocephalus: morphology and volumetry. AJNR Am J Neuroradiol 19: , Klinge P, Hellström P, Tans J, Wikkelsø C: One-year outcome in the European multicentre study on inph. Acta Neurol Scand 126: , LeMay M, New PF: Radiological diagnosis of occult normalpressure hydrocephalus. Radiology 96: , Malm J, Kristensen B, Stegmayr B, Fagerlund M, Koskinen LO: Three-year survival and functional outcome of patients with idiopathic adult hydrocephalus syndrome. Neurology 55: ,

7 J. Virhammar et al. 18. Marmarou A, Young HF, Aygok GA, Sawauchi S, Tsuji O, Yama moto T, et al: Diagnosis and management of idiopathic nor mal-pressure hydrocephalus: a prospective study in 151 patients. J Neurosurg 102: , Mori E, Ishikawa M, Kato T, Kazui H, Miyake H, Miyajima M, et al: Guidelines for management of idiopathic normal pressure hydrocephalus: second edition. Neurol Med Chir (Tokyo) 52: , Relkin N, Marmarou A, Klinge P, Bergsneider M, Black PM: Diagnosing idiopathic normal-pressure hydrocephalus. Neurosurgery 57 (3 Suppl):S4 S16, ii v, Sasaki M, Honda S, Yuasa T, Iwamura A, Shibata E, Ohba H: Narrow CSF space at high convexity and high midline areas in idiopathic normal pressure hydrocephalus detected by axial and coronal MRI. Neuroradiology 50: , Savolainen S, Hurskainen H, Paljärvi L, Alafuzoff I, Vapalahti M: Five-year outcome of normal pressure hydrocephalus with or without a shunt: predictive value of the clinical signs, neuropsychological evaluation and infusion test. Acta Neurochir (Wien) 144: , Sjaastad O, Nordvik A: The corpus callosal angle in the diagnosis of cerebral ventricular enlargement. Acta Neurol Scand 49: , Steffen TM, Hacker TA, Mollinger L: Age- and gender-related test performance in community-dwelling elderly people: Six- Minute Walk Test, Berg Balance Scale, Timed Up & Go Test, and gait speeds. Phys Ther 82: , Tarnaris A, Kitchen ND, Watkins LD: Noninvasive biomarkers in normal pressure hydrocephalus: evidence for the role of neuroimaging. A review. J Neurosurg 110: , Virhammar J, Cesarini KG, Laurell K: The CSF tap test in normal pressure hydrocephalus: evaluation time, reliability and the influence of pain. Eur J Neurol 19: , Walter C, Hertel F, Naumann E, Mörsdorf M: Alteration of cerebral perfusion in patients with idiopathic normal pressure hydrocephalus measured by 3D perfusion weighted magnetic resonance imaging. J Neurol 252: , Wikkelsø C, Hellström P, Klinge PM, Tans JT: The European inph Multicentre Study on the predictive values of resistance to CSF outflow and the CSF Tap Test in patients with idiopathic normal pressure hydrocephalus. J Neurol Neurosurg Psychiatry 84: , Wilson RK, Williams MA: The role of the neurologist in the longitudinal management of normal pressure hydrocephalus. Neurologist 16: , 2010 Manuscript submitted March 21, Accepted August 20, Results from this work were presented at two conferences last year: Hydrocephalus 2012 in Kyoto, Japan, October 19 22, 2012, and ESMRMB 2012 in Lisbon, Portugal, October 4 6, Please include this information when citing this paper: published online September 27, 2013; DOI: / JNS Address correspondence to: Johan Virhammar, M.D., Department of Neuroscience, Neurology, Uppsala University, Akademiska Sjukhuset, ing 85, Uppsala, Sweden. johan.virhammar@ neuro.uu.se. 184 J Neurosurg / Volume 120 / January 2014