- Email: [email protected]
CSF tap test — Obsolete or appropriate test for predicting shunt responsiveness? A systemic review
Journal of the Neurological Sciences 362 (2016) 78–84
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Review article
CSF tap test — Obsolete or appropriate test for predicting shunt responsiveness? A systemic review Mario Mihalj a,⁎, Krešimir Dolić b, Krešimir Kolić b, Vlatko Ledenko c a b c
Department of Neurology, University Hospital Split, Croatia Clinical Department of Interventional and Diagnostic Radiology, University Hospital Split, Croatia Department of Neurosurgery, University Hospital Split, Croatia
a r t i c l e
i n f o
Article history: Received 24 April 2015 Received in revised form 13 January 2016 Accepted 18 January 2016 Available online 22 January 2016 Keywords: Normal pressure hydrocephalus Idiopathic normotensive hydrocephalus Shunt operation CSF tap test Predictive value Validity
a b s t r a c t Objectives: There is no accurate test for diagnosing normal pressure hydrocephalus or for screening for patients who will benefit from shunt surgery. Additional tests, such as cerebrospinal fluid tap test (CSF-TT), are often used in practice to provide further predictive value in detecting suitable patients for shunting. We performed a systematic review of the literature to evaluate the CSF-TT's effect on the outcome of main symptoms and on validity parameters in screening patients suitable for shunting. Methods: In February 2015 we searched electronic databases from their inception to the current date, using the following key words: normal pressure hydrocephalus, idiopathic normotensive hydrocephalus, shunt operation, CSF tap test, predictive value, validity. The search retrieved 8 articles explicitly addressing the topic. Results: There was a very high positive predictive value of CSF-TT: 92% (range from 73% to 100%) but a low negative predictive value: 37% (18%–50%). Also, the CSF-TT has high specificity: 75% (33%–100%) but average sensitivity: 58% (26%–87%). The overall accuracy of the test was 62% (45%–83%). Conclusions: This systematic review did not provide unambiguous validity of the CSF-TT in the screening of patients for shunting. The validity of the CSF-TT is good for patient inclusion for shunting due to the fact that the positive response to the test is very reliable. Unfortunately, the negative response to the test does not reliably make these patients ineligible for shunting. Further studies are needed to improve and standardize the methodology in order to optimize the detection power of the test. © 2016 Elsevier B.V. All rights reserved.
Contents 1. 2.
3.
4.
Introduction . . . . . . . . . . . . . . . . Methods . . . . . . . . . . . . . . . . . . 2.1. Study design . . . . . . . . . . . . . 2.1.1. Search strategy. . . . . . . . 2.2. Study selection . . . . . . . . . . . . 2.2.1. Inclusion and exclusion criteria 2.2.2. Data extraction. . . . . . . . 2.3. Assessment of methodological quality . 2.3.1. Diagnostic accuracy measures . Results . . . . . . . . . . . . . . . . . . . 3.1. Study selection . . . . . . . . . . . . 3.1.1. Study characteristics . . . . . 3.2. Diagnostic accuracy measures . . . . . 3.2.1. Methodological quality . . . . 3.2.2. Characteristics of studies . . . Discussion . . . . . . . . . . . . . . . . . 4.1. Limitations. . . . . . . . . . . . . . 4.2. Suggestions for future research . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
⁎ Corresponding author at: Department of Neurology, University Hospital Split, Spinčićeva 1, 21000 Split, Croatia. E-mail address: [email protected] (M. Mihalj).
http://dx.doi.org/10.1016/j.jns.2016.01.028 0022-510X/© 2016 Elsevier B.V. All rights reserved.
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
79 79 79 79 79 79 80 80 81 81 81 81 81 81 81 82 83 83
M. Mihalj et al. / Journal of the Neurological Sciences 362 (2016) 78–84
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Declaration of interests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction In 1965, Hakim and Adams described the syndrome of gait disturbance, cognitive deterioration and urinary incontinence — a clinical triad associated with ventricular enlargement disproportional to any sulcal enlargement (distinguishing it from atrophy), in the absence of elevated cerebrospinal fluid (CSF) pressure during lumbar puncture [1,2]. Normal pressure hydrocephalus (NPH) in patients without known precipitants is termed primary or idiopathic iNPH (iNPH). When it occurs after other diseases, such as meningitis, traumatic brain injury, subarachnoid hemorrhage, cerebral infarction, this syndrome is called secondary NPH [3,4]. Some authors point to a hereditary predisposition towards iNPH [5,6]. There is variation in the clinical presentation, severity and progression of these symptoms, and it is not necessary for the entire triad to be present in order to consider diagnosing iNPH. Gait and balance impairment appear either before or concurrently with urinary incontinence or the onset of dementia. Symptoms are developed insidiously, and generally occur between the sixth and eighth decade of life [7,8]. Gait disturbances are the first signs of iNPH, and described as apraxic, glue-footed, magnetic, bradykinetic, and shuffling gait [9,10,11]. They are often misinterpreted as symptoms of Parkinson's disease [12]. Urinary incontinence usually follows gait abnormalities and almost always includes urinary urgency [13,14]. Dementia is rarely the first and foremost symptom of NPH, although it is often present [9, 13]. The impairment is mainly cognitive and subcortical in type, characterized by inattention, delay in responding and remembering, lack of spontaneity but without cognitive decline as in cortical dementia [13, 15]. Although NPH is commonly referred to as a treatable form of dementia, cognitive deficits and memory loss are the symptoms less likely to respond to shunting [16,17,18]. Neuroimaging with CT or MRI is an essential part of the evaluation of patients with suspected NPH/iNPH and ventricular enlargement is necessary to establish the diagnosis of NPH for patients with appropriate symptoms. A frontal horn ratio (Evans' index), defined as the maximal frontal horn ventricular width divided by the transverse inner diameter of the skull, indicates ventriculomegaly if it is 0.3 or greater [10,18]. Diagnosis based on clinical and radiological signs alone can be problematic and additional diagnostic testing may be required to determine which patients could benefit from shunting. A transient, favorable clinical effect in three NPH patients after removal of 15 ml of fluid was first described by Adams et al. [2]. CSF-TT was later modified by Wikkelso et al. by introducing the removal of larger quantities of fluid (40–50 ml) and quantitative testing of the main symptoms [19,20]. Although inexpensive, readily available and safe (without serious complications), there are controversial opinions about CSF-TT's detection power and validity in screening patients suitable for shunting. We searched the Cochrane Database of Systematic Reviews (CDSR), but no systematic review on this topic was found. Therefore, a systematic review was conducted to determine clinical impact and validity of CSF-TT in the prediction of response to shunting. 2. Methods 2.1. Study design The methods of this systematic review were decided a priori and adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [21]. The PRISMA statement
79
83 83 84
includes a 27-item checklist designed to improve reporting of systematic reviews and meta-analyses. 2.1.1. Search strategy Several online databases (PubMed, OvidMedline, Current Contents, Science Direct, Ebsco, and Web of Science) were searched from their inception to February 30, 2015. Authors (M.M., K.D., K.K.,V.L.) typed keywords independently from one another using the Boolean operator OR & AND and searched the online databases. The search strategy was as follows: ((normal pressure hydrocephalus) OR (normotensive hydrocephalus) OR (idiopathic normal pressure hydrocephalus)) AND (shunt operation) AND ((CSF tap test) OR (spinal tap test)) AND predictive value. A total of 2898 published articles were searched by abstract or full text. Retrospective and prospective studies were also searched. There were no language restrictions. Any published article that explicitly addressed CSF-TT was included, particularly if the article provided numerical data that could be included in a systematic review. A meta-analysis could not be performed because most of the articles were not designed as a randomized controlled trial. The search strategy is shown in Fig. 1. A complementary manual search of reference lists and personal resources was also performed to identify any relevant articles missed in the electronic searches. Finally, we searched for associated publications of retrieved articles to obtain the most complete and up-to-date study results. Papers that did not present clinical data were excluded, and any duplicate presentation of clinical data was identified. 2.2. Study selection Two authors (M.M. and K.D.) reviewed the electronic database search results (title and abstract) independently. Any titles and abstracts that appeared to meet inclusion criteria were selected for full text review. The reference lists of the identified studies were reviewed to discover additional potentially eligible studies. Unpublished data and conference proceedings were excluded from this review. Abstracts were excluded when both investigators agreed they were not relevant. Any disagreements were resolved within discussion between the two authors. The same two authors independently conducted the full text review of the retrieved articles in regard to the inclusion and exclusion criteria. Any disagreements were resolved by discussion, with a third and fourth author (K.K., V.L.) consulted if resolution was not achieved, to produce the final articles for inclusion. A data extraction form was prepared, with two authors (M.M., K.D.) independently extracting data from the selected studies. Authors (M.M., K.K.) reviewed the completed form for accuracy, with any disagreements resolved by the fourth author (V.L.). 2.2.1. Inclusion and exclusion criteria In our analysis, we included studies that met the following prespecified criteria: a diagnosis of NPH (iNPH) based on clinical examination and neuroimaging (CT/MRI), outcome data for at least for two of the three main symptoms (gait, cognition, continence) after CSF-TT and shunting, follow up and reevaluation of outcomes (at least once and not earlier than three months after shunting). NPH (iNPH) was defined as a symmetrical quadriventricular enlargement (Evans' Index ≥ 0.3) without clinically significant cortical and parenchymal lesions (atrophy, infarcts) with free communication between the ventricular system and the subarachnoid space (“communicating”
80
M. Mihalj et al. / Journal of the Neurological Sciences 362 (2016) 78–84
Fig. 1. Flow diagram for identification of relevant studies.
hydrocephalus) verified by CT/MR neuroimaging. Also, for the diagnosis to be made we required the presence of at least two symptoms of the clinical triad (gait was mandatory) Studies were not included if they did not provide data on the main symptoms and parameters of validity of the CSF-TT or raw data on outcomes of the main symptoms after CSF-TT and shunting from which it was possible to assess the parameters of validity. Furthermore, studies which did not describe (quantitatively and qualitatively) the reasons why patients were lost to follow-up were not included. Also, studies where shunting outcomes were not taken as reference (“gold standard”) for the definitive diagnosis and treatment of NPH were excluded.
symptoms), inclusion and exclusion criteria, the setting for patient recruitment - whether recruitment was consecutive and/or data collection was performed prospectively, data on the diagnosis of NPH (clinical, neuroimaging — CT/MRI), data on the outcomes for the main symptoms (gait, cognition, continence), number of patients lost to follow-up. We also collected: description and timing of outcome measures, data on the amount of removed CSF, CSF opening pressure, and complications after CSF-TT.
2.2.2. Data extraction The following data were extracted: author, year of publication, characteristics of the study population (age, gender, prevalence of the main
The methodological quality of the studies was assessed by the 2 reviewers (M.M., K.D.) using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool [22].
2.3. Assessment of methodological quality
M. Mihalj et al. / Journal of the Neurological Sciences 362 (2016) 78–84
The reviewers were blind to each other's assessment and scored the criteria items with “yes” or “no” when studies satisfied or failed to meet the criteria, respectively, and “unclear” when information was lacking to decide whether the study satisfied or met that specific item. In the case of disagreement, the 2 reviewers tried to reach a consensus on each criterion, and in case of persisting disagreement, a third or fourth reviewer (K.K. and V.L.) was involved. 2.3.1. Diagnostic accuracy measures For quantitative assessment, statistical measures of predictive validity (sensitivity, specificity, positive and negative predictive values, accuracy) from each included study were extracted or calculated from cumulative (2 × 2) tables completed by one of the authors (M.M.) and confirmed by another author (K.D.). 3. Results 3.1. Study selection Fig. 1 depicts the flow of articles through the review process. Eight articles, evaluating a total of 482 patients, met the inclusion criteria for this review. Seven of these articles were identified through the electronic database search [20,24–29] while the remaining article [23] was identified by reference searching. The results were heterogeneous and varied from very good [20,25, 28] to disappointingly limited power of the CSF-TT [24,26,29] in predicting the response to shunting. 3.1.1. Study characteristics Tables 1 and 2 present the characteristics of the studies included in the review. 3.2. Diagnostic accuracy measures Table 3 presents the parameters of the CSF-TT's validity in the included studies. 3.2.1. Methodological quality Fig. 2 presents the QUADAS 2 assessment results for each included study. Five studies were found to have low risk of bias [20,25,26,28, 29], and three studies were deemed to be at risk of bias [23,24,27]. Six studies were deemed to have low concern regarding applicability [20,
81
25,26,27,28,29], whereas two studies had concerns regarding applicability [23,24].
3.2.2. Characteristics of studies Haan and Thomeer examined the effect of the CSF-TT and external lumbar drainage (ELD) on main NPH symptoms in 32 iNPH patients. The CSF-TT had a low sensitivity (43%) and negative predictive value (42%) but specificity and positive predictive value were very high, 100% [23]. Wikkelsø and colleagues [20] studied the predictive value in 27 patients (21 NPH/6 iNPH), 15 patients were included from the previous study by the same authors and reported good sensitivity (68%), lower negative predictive value (45%) but positive predictive value and specificity was very high-100%. Malm et al. [24] compared the results after CSF-TT and CSF flow conductance with the outcomes after shunt surgery in 35 iNPH cases and showed low negative predictive value (23%) and specificity (33%) while the positive predictive value was 73% and sensitivity was 62%. Damasceno et al. [25] reported very high values for positive predictive value (93%) and sensitivity (87%), specificity was 67% but the negative predictive value was 50%. Kahlon et al. [27] compared results between the lumbar infusion test (LIT) and CSF-TT with outcomes after shunting in 47 (35 iNPH/12 NPH) patients. The total number of false negative values was unknown because NPH patients, with negative results in both LIT and CSF-TT (21/47), were not operated on. Therefore, the exact values for sensitivity and specificity could not be calculated. For those who were operated on, parameters of validity were as follows: positive predictive value — 94%, specificity — 88.9%, negative predictive value — 26.7% and sensitivity — 42%. Walchenbach et al. studied the effect of external lumbar drainage (ELD) and CSF-TT in 47 patients and showed that CSF- TT had low sensitivity (26%) and negative predictive value (32%). On the contrary, it had very high specificity and positive predictive value - 100% [26]. Ishikawa and colleagues analyzed the results of the effect of the tap test from the SINPHONI (Study of Idiopathic Normal Pressure Hydrocephalus on Neurological Improvement). The main conclusion was that the sensitivity and specificity of the CSF-TT were optimum when improvement on any iNPH grading scale was combined with CSF opening pressure ≥ 15 cm H20 [28]. Therefore sensitivity rose from 71% to 83% while specificity remained at 65%. The positive and negative predictive values were 89% and 50%, respectively [28]. In the Eu-iNPH study by Wikkelsø et al. a very low negative predictive value (18%) was obtained due to the large number of false negatives. The test sensitivity was 52%. The
Table 1 Demographic data, diagnostic methods, follow - up time and prevalence of main NPH symptoms. Study
Age (years) Mean ± (SD) or range
Gender (F/M)
Number of patients/lost iNPH (NPH) diagnose to follow-upa
Follow-up after shunt (months)
Cognition Gait disturbance impairment (%) (%)
Incontinence (%)
Wikkelsø et al. [29] Ishikawa et al. [28] Walchenbach et al. [26] Kahlon et al. [27] Damasceno et al. [25] Malm et al. [24] Haan and Thomeer [23] Wikkelsø et al. [20]
71 (30–87)
69/73
142/27
Clinically & radiologically (CT/MR)
1, 3, 6, 12
Missing data
Missing data
75 (71–78)
42/58
100
Clinically & radiologically (MR)
3, 6, 12
Missing data 91%
80%
60%
77 (55–87)
14/37
48/1
Clinically & radiologically (MR)
2, 6, 12
Missing data
Missing data
72 ± 9
39/29
81/13
Clinically & radiologically (CT)
6 (4,5)
Missing data 91%
76%
60%
18
Clinically & radiologically (CT)
3, 6,12
100%
83%
83%
a
66.5 ± 9.6
6/12
71.7 ± 5.2
10/27
35/2
Clinically & radiologically (CT)
3
100%
49%
59%
8/24
32/6
Clinically & radiologically (CT)
3, 6, 12, 24
100%
90%
78%
14/13
27/3
Clinically & radiologically (CT/radionuclide cisternography)
3–6
89%
74%
81%
70.5 (58–80)
63 (40–81)
Number of patients who underwent CSF-TT/lost to follow-up.
82
M. Mihalj et al. / Journal of the Neurological Sciences 362 (2016) 78–84
Table 2 Data on the amount of CSF removed, opening pressure, outcome measures, time of evaluation and possible complications. Author
Amount of CSF (ml)
Opening CSF pressurea
Gait (Outcome measuresb)
Cognition (Outcome measuresb)
Continence (Outcome measuresb)
Complications
Wikkelsø et al. [20]
50
18 m walking test (3×), 2 h after CSF TT
Identical forms test, Bingleys memory test, reaction time test, 2 h after CSF TT
Missing data
No complication
Malm et al. [24] Damasceno et al. [25] Kahlon et al. [27] Walchenbach et al. [26] Ischikawa et al. [28] Wikkelsø et al. [29]
Different
16.32 ± 6.8 (5.44–29.4) cm H20 0 cm H20
2 headaches
b18 cm H20
50
Unknown
40
Unknown
30
12 (9–14) cm H20 and ≥15 13.6 (2.7–24.5)
MMSE, series of neuropsychological tests (Bingley's Memory, Digit Span, Peg Bord) Test of visuo-motor speed, visuo-constructive skills, memory test, 2–8 h after CSF TT Reaction time test, identical forms test, memory test-Bingley's charts, 1–3 h after CSF TT MMSE, series of neuropsychological tests, 6–8 h after CSF TT iNPHGS (0–4), MMSE, 7 days after CSF-TT
Missing data
50
25 m walking test (3×), Barthel index of ADL, 3 to 4 h after CSFTT 18 m walking test (4×), 2–8 h after CSF TT 18 m walking test (3×), 1–3 h after CSF-TT An ordinal (0–5) rating scale, 6–8 h after CSF TT iNPHGS(0–4), mRS, TUG, 1–2 days after CSF TT mRS, iNPH rating scale (1–8), 3 h after CSF TT
a b
50
cm H20
MMSE, three neuropsychological tests, 3 h after CSF TT
Missing data
No complication Missing data No complication An ordinal (3 point) No rating scale complication iNPHGS (0–4) No 7 days after CSF TT complication iNPH rating scale No (1–6), 3 h after CSF TT complication
Presented as mean value or range. Tests for evaluating outcomes before and after CSF-TT and shunting.
positive predictive value and specificity were 88% and 59%, respectively [29].
4. Discussion Our systematic review of eight prospective cohort studies, involving 482 cases of normal pressure hydrocephalus, demonstrated a questionable validity of the CSF-TT in predicting response to shunt surgery. This was particularly true in regard to the negative response to the test. The positive predictive value of CSF-TT was very high across the majority of analyzed studies, ranging between 73% [24] and 100% [20,23,26]. It was due to the fact that most patients - 80% (range 73%–86%) improved after shunting, which gave a high positive predictive value simply by chance. Similarly, due to the low percentage of false positives [20,23,26], specificity was also high (75%), ranging between 33% [24] and 100% [20,23,26]. This highlights the utility of CSF-TT as a rule-in test for shunting when positive. Only the study by Malm et al. showed a low specificity (33%) (many false positives), but also a low negative predictive value — 23% (many false negatives) [24]. The main reason for this finding was the removal of different amounts of CSF in different patients depending on the volume and elasticity of the system [24]. In contrast, the sensitivity was extremely heterogeneous, ranging between 26% in the study by Walchenbach et al. [26] and 87% in the study by Damasceno et al. [25]. Also, the negative predictive value was low, ranging between 18% in the study by Wikkelsø et al. [29] and 50% in the studies by Damasceno and Ishikawa et al. [25,28]. The accuracy was 62%, ranging between 45% in the study by Walchenbach et al. [26] and 83% in the study by Damasceno et al. [25].
A key source of the heterogeneity of the results for these parameters of validity of the CSF-TT is the lack of standardization of methodology in assessing the main symptoms of NPH in the analyzed studies. The amount of CSF removed varied from 30 ml, in the study by Ishikawa et al. [28] and 40 ml in the study by Walchenbach et al. [26] to 50 ml in other studies [20,25,27,29]. In the study by Malm et al. [24] different amounts of CSF were removed, and the study of Haan and Thomeer did not state this data [23]. It seems that the amount of CSF removed affects certain parameters of the test validity; higher sensitivity and negative predictive values were observed in the studies where greater amounts of cerebrospinal fluid were removed [20,25] than in the studies of Malm [24], and Walchenbach [26]. This was also highlighted by Kilic et al. [30] and Damasceno [31]. They introduced the removal of larger amounts of fluid by repeating CSF-TT (RTT) during two or three consecutive days, removing 30– 50 ml/d, to the total of 100–120 ml. The authors reported improved sensitivity, negative and positive predictive value, actually empowering the detection of responders for shunting. Since the test has no serious complications compared with shunt surgery, some authors recommend repeating the test [32]. The positive response to RTT is recommended as one of the predictors of successful response to shunting, as recently published by the American Academy of Neurology (AAN) practice guideline for predictors of shunting effectiveness and utility of shunting in iNPH [33]. Conversely, a disappointingly low negative predictive value (18%) is reported in the Eu-iNPH study by Wikkelsø et al., despite the volume of the removed fluid. The authors concluded that the power of CSF-TT to predict a good outcome of shunting was disappointingly limited [29]. Also, the heterogeneity between analyzed studies exists in terms of
Table 3 Parameters of validity of CSF-TT in assessing suitable patients for shunting. Author
Region
Total number of patients, design of study
PPV (%)
NPV (%)
SN (%)
SP (%)
AC (%)
Wikkelsø et al. [20] Haan and Thomeer [23] Malm et al. [24] Damasceno et al. [25] Kahlon et al. [27] Walchenbach et al. [26] Ishikawa et al. [28] Wikkelsø et al. [29]
Europe, Sweden Europe, The Netherlands Europe, Sweden South America, Brazil Europe, Sweden Europe, The Netherlands Asia, Japan Europe, 9 countries
27, prospective 32, prospective 35, prospective 18, prospective 81, prospective 47, prospective, 3 centers 100, prospective, 26 centers 142, single blinded, prospective, 13 centers
100 100 73 93 94 100 89 88
45 42 23 50 Unknowna 32 50 18
68 43 62 87 Unknowna 26 71 52
100 100 33 67 Unknowna 100 65 59
75 54 54 83 – 45 70 53
Legend: PPV = positive predictive value; NPV = negative predictive value; SN = sensitivity, SP = specificity; AC = accuracy. a Unknown number of negatives (false and true).
M. Mihalj et al. / Journal of the Neurological Sciences 362 (2016) 78–84
83
Fig. 2. Graphical presentation for QUADAS-2 results in diagnostic accuracy.
the CSF opening pressure, ranging from 0 cm H2O in the study of Malm et al. [24] to 29.4 cm H2O in the studies by Wikkelsø et al. [20]. In the two studies, data were not presented on the CSF opening pressure [26,27]. In the studies by Damasceno [25] and Ishikawa [28], where the opening pressures were set higher, higher negative predictive values (50%) and sensitivity were obtained (87% and 83%, respectively). Although the higher CSF opening pressure could indicate the possibility of secondary forms of NPH, Ishikawa and colleagues [28] pointed to the fact that there was a drop in false negatives when the CSF opening pressure was set at ≥ 15 cm H2O. Therefore, the sensitivity increased from 71% to 83%, while the specificity remained the same (65%). In clinical practice, fewer shunt responders would be missed. There was tremendous heterogeneity in the labeling of “improved” gait among analyzed studies. In the study by Haan and Thomeer, the gait improvement was too strictly categorized as mobile vs. immobile [23], while in the study by Walchenbach et al. [26] the scoring was more lenient (e.g., improvement for at least one point on the fivepoint gait score was considered markedly positive). Depending on the stipulated cut-off level for „significant improvement“, different parameters were obtained for the validity of CSF-TT among analyzed studies. For example, applying more stringent criteria or limits for the labeling of “improved” gait increases the number of false negatives and consequently lowers the negative predictive value and sensitivity, but also the accuracy of the test, as highlighted in recent studies [28,29]. Only in two recent studies, the newly developed iNPH scale was used, reducing the discrepancy in the assessment of outcomes, especially in the gait domain [28,29]. It is known that NPH patients have daily fluctuations in the performance of cognitive tasks and movements [8,9,10]. The assessment time of the outcomes of the main symptoms, particularly in the gait and neuropsychological domains, was significantly different among the analyzed studies: 1–4h [20,24,27,29]; 2–8 h [25,26]; and 1–2 days, [28] after performing CSF-TT. Only in the study by Ishikawa et al. [28] neuropsychological and continence tests were performed seven days after the tap test. According to previously published studies, the fastest and most obvious improvements were observed in the gait domain after CSF-TT and shunting, but the severity and duration of the gait disturbance (and other symptoms) were negatively associated with favorable outcomes after CSFTT and shunting [23–25,29,33,34]. Virhammar et al. highlighted that better results in the gait domain were obtained when the assessment was carried out several hours after CSF-TT, repeatedly during 24 h. It is particularly important to adequately prevent or treat post-lumbar puncture pain, which has a negative effect on outcomes [35]. In most studies, the Mini Mental State Examination (MMSE) was used for the assessment of outcomes, with several other similar neuropsychological tests. MMSE is recommended in the early assessment, because it shows more prominent changes within the first day after CSF-
TT than other neuropsychological tests [36]. Unfortunately, only two studies recorded improvement in the neuropsychological domain after CSF-TT and positive correlation with improvement after shunting [20, 27]. There was no recovery in the continence domain after CSF-TT in any of the analyzed studies, [20,23–29], which was expected due to the more difficult and long-term recovery of this symptom. Also, there were no serious complications during and after the test. 4.1. Limitations The lack of a standardized methodology was one of the main challenges of this review, as evidenced by clinical and statistical heterogeneity. The general quality of the included articles was moderate. Two studies were considered to be at high risk of bias, one within the index test domain because of its interpretation and another study was within the flow and timing domain because the same reference standard was not used for all patients and not all samples included in the study had been accounted for in the analysis (the number of negatives, especially false negatives was unknown). One study was judged as “unclear” in the index test domain because of insufficient data. In two studies the applicability was considered to be of high concern with regard to execution and interpretation of the index test. Also, one of the shortcomings of the analyzed studies was the lack of data on the extent and severity of the impairment for each of the main symptoms at baseline, which certainly affected the results of the outcomes. All of the above may have resulted in misclassification and estimated parameters of validity may have been biased. 4.2. Suggestions for future research In future, study methodology, including the measuring of the amount of CSF removed, opening pressure, cut-off level for labeling “improved” and timing of outcome assessment of the main symptoms after the tap test, should be standardized in order to reduce the risk of bias and enable better comparability. 5. Conclusions Inexpensive, readily available, and without serious side-effects, CSFTT is still the first choice of additional tests in the detection of suitable patients for shunt surgery. Due to lower reliability of the negative findings (low negative predictive value and sensitivity), the test should not be used for excluding patients from shunt surgery because many shunt responders could be missed. Declaration of interests The authors declare that they have no conflict of interest.
84
M. Mihalj et al. / Journal of the Neurological Sciences 362 (2016) 78–84
References [1] S. Hakim, R.D. Adams, The special clinical problem of symptomatic hydrocephalus with normal cerebrospinal fluid pressure: observations on cerebrospinal fluid hydrodynamics, J. Neurol. Sci. 2 (1965) 307–327. [2] R.D. Adams, C.M. Fisher, S. Hakim, R.G. Ojemann, W.H. Sweet, Symptomatic occult hydrocephalus with “normal” cerebrospinal-fluid pressure: a treatable syndrome, N. Engl. J. Med. 273 (1965) 117–126. [3] A.O. Hebb, M.D. Cusimano, Idiopathic normal pressure hydrocephalus: a systematic review of diagnosis and outcome, Neurosurgery 49 (2001) 1166–1184. [4] B. Beyerl, P.M. Black, Posttraumatic hydrocephalus, Neurosurgery 15 (1984) 257–261. [5] M.D. Cusimano, D. Rewilak, D.T. Stuss, J.C. Barrera-Martinez, F. Salehi, M. Freedman, Normal-pressure hydrocephalus: is there a genetic predisposition? Can J. Neurol. Sci. 38 (2011) 274–281. [6] A. McGirr, M.D. Cusimano, Familial aggregation of idiopathic normal pressure hydrocephalus: novel familial case and a family study of the NPH triad in an iNPH patient cohort, Neurol. Sci. 321 (2012) 82–88. [7] C.M. Fisher, The clinical picture in occult hydrocephalus, Clin. Neurosurg. 24 (1977) 270–284. [8] M.J. McGirt, G. Woodworth, A.L. Coon, G. Thomas, M.A. Williams, D. Rigamonti, Diagnosis, treatment, and analysis of long-term outcomes in idiopathic normal-pressure hydrocephalus, Neurosurgery 57 (2005) 699–705. [9] P.L. Gleason, P.M. Black, M. Matsumae, The neurobiology of normal pressure hydrocephalus, Neurosurg. Clin. North Am. 4 (1993) 667–675. [10] N. Relkin, A. Marmarou, P. Klinge, M. Bergsneider, P.M. Black, Diagnosing idiopathic normalpressure hydrocephalus, Neurosurgery 57 (2005) S4–S16. [11] C.M. Fisher, Hydrocephalus as a cause of disturbances of gait in the elderly, Neurology 32 (1982) 1358–1363. [12] M. Mihalj, M. Titlić, A. Marović, S. Srdelić-Mihalj, Gait apraxia, Bratisl. Lek. Listy 111 (2010) 101–102. [13] J.A. Vanneste, Diagnosis and management of normalpressure hydrocephalus, J. Neurol. 247 (2000) 5–14. [14] A. Larsson, C. Wikkelsö, M. Bilting, H. Stephensen, Clinical parameters in 74 consecutive patients shunt operated for normal pressure hydrocephalus, Acta Neurol. Scand. 84 (1991) 475–482. [15] P. Bret, J. Guyotat, J. Chazal, Is normal pressure hydrocephalus a valid concept in 2002? A reappraisal in five questions and proposal for a new designation of the syndrome as “chronic hydrocephalus”, J. Neurol. Neurosurg. Psychiatry 73 (2002) 9–12. [16] J.K. Krauss, M. Faist, M. Schubert, J.J. Borremans, C.H. Lücking, W. Berger, Evaluation of gait in normal pressure hydrocephalus before and after shunting, Adv. Neurol. 87 (2001) 301–310. [17] P.M. Black, Idiopathic normal-pressure hydrocephalus. Results of shunting in 62 patients, J. Neurosurg. 52 (1980) 371–377. [18] R.C. Petersen, B. Mokri, E.R. Laws Jr., Surgical treatment of idiopathic hydrocephalus in elderly patients, Neurology 35 (1985) 307–311. [19] C. Wikkelsø, H. Andersson, C. Blomstrand, G. Lindqvist, The clinical effect of lumbar puncture in normal pressure hydrocephalus, J. Neurol. Neurosurg. Psychiatry 45 (1982) 64–69.
[20] C. Wikkelsø, H. Andersson, C. Blomstrand, G. Lindqvist, P. Svendsen, Normal pressure hydrocephalus: predictive value of the cerebrospinal fluid tap-test, Acta Neurol. Scand. 73 (1986) 566–573. [21] D. Moher, A. Liberati, J. Tetzlaff, D. Altman, PRISMA GROUP, Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement, PLoS Med. 6 (2009). http://dx.doi.org/10.1371/journal.pmed.1000097. [22] P. Whiting, A. Rutjes, M. Westwood, S. Mallett, J. Deeks, J. Reitsma, et al., QUADAS-2 Group, QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies, Ann. Intern. Med. 155 (2011) 529–536. [23] J. Haan, R.T. Thomeer, Predictive value of temporary external lumbar drainage in normal pressure hydrocephalus, Neurosurgery 22 (1988) 388–391. [24] M. Malm, B. Kristensen, T. Karlsson, M. Fagerlund, J. Elfverson, J. Ekstedt, The predictive value of cerebrospinal fluid dynamic tests in patients with the idiopathic adult hydrocephalus syndrome, Arch. Neurol. 52 (1995) 783–789. [25] B.P. Damasceno, E.F. Carelli, D.C. Honorato, J.J. Facure, The predictive value of cerebrospinal fluid tap-test in normal pressure hydrocephalus, Arq. Neuropsiquiatr. 55 (1997) 179–185. [26] R. Walchenbach, E. Geiger, R.T. Thomeer, J.A. Vanneste, The value of temporary external lumbar CSF drainage in predicting the outcome of shunting on normal pressure hydrocephalus, J. Neurol. Neurosurg. Psychiatry 72 (2002) 503–506. [27] B. Kahlon, G. Sundbärg, S. Rehncrona, Comparison between the lumbar infusion and CSF tap tests to predict outcome after shunt surgery in suspected normal pressure hydrocephalus, J. Neurol. Neurosurg. Psychiatry 73 (2002) 721–726. [28] M. Ishikawa, M. Hashimoto, E. Mori, N. Kuwana, H. Kazui, The value of the cerebrospinal fluid tap test for predicting shunt effectiveness in idiopathic normal pressure hydrocephalus, Fluids Barriers CNS 13:9 (1) (2012) 1. [29] C. Wikkelsø, P. Hellström, P.M. Klinge, J.T. Tans, European iNPH Multicentre Study Group, 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 (2013) 562–568. [30] K. Kilic, A. Czorny, J. Auque, Z. Berkman, Predicting the outcome of shunt in normal pressure hydrocephalus, J. Clin. Neurosci. 14 (2007) 729–736. [31] B.P. Damasceno, Normal pressure hydrocephalus. Diagnostic and predictive evaluation, Dement. Neuropsychol. 3 (2009) 8–15. [32] T.S. Lim, S.W. Yong, S.Y. Moon, Repetitive lumbar punctures as treatment for normal pressure hydrocephalus, Eur. Neurol. 62 (2009) 293–297. [33] J.J. Halperin, R. Kurlan, J.M. Schwalb, M.D. Cusimano, G. Gronseth, D. Gloss, Practice guideline: idiopathic normal pressure hydrocephalus: response to shunting and predictors of response: report of the guideline development, dissemination, and implementation subcommittee of the American academy of neurology, Neurology 85 (2015) 2063–2071. [34] U. Meier, J. Lemcke, U. Neumann, Predictors of outcome in patients with normalpressure hydrocephalus, Acta Neurochir. 96 (2006) 352–357. [35] J. Virhammar, K.G. Cesarini, K. Laurell, The CSF tap test in normal pressure hydrocephalus: evaluation time, reliability and the influence of pain, Eur. J. Neurol. 19 (2012) 271–276. [36] H. Schmidt, J. Elster, I. Eckert, J. Wiefek, W. Paulus, N. von Steinbuechel, et al., Cognitive functions after spinal tap in patients with normal pressure hydrocephalus, J. Neurol. 261 (2014) 2344–2350.
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Review article
CSF tap test — Obsolete or appropriate test for predicting shunt responsiveness? A systemic review Mario Mihalj a,⁎, Krešimir Dolić b, Krešimir Kolić b, Vlatko Ledenko c a b c
Department of Neurology, University Hospital Split, Croatia Clinical Department of Interventional and Diagnostic Radiology, University Hospital Split, Croatia Department of Neurosurgery, University Hospital Split, Croatia
a r t i c l e
i n f o
Article history: Received 24 April 2015 Received in revised form 13 January 2016 Accepted 18 January 2016 Available online 22 January 2016 Keywords: Normal pressure hydrocephalus Idiopathic normotensive hydrocephalus Shunt operation CSF tap test Predictive value Validity
a b s t r a c t Objectives: There is no accurate test for diagnosing normal pressure hydrocephalus or for screening for patients who will benefit from shunt surgery. Additional tests, such as cerebrospinal fluid tap test (CSF-TT), are often used in practice to provide further predictive value in detecting suitable patients for shunting. We performed a systematic review of the literature to evaluate the CSF-TT's effect on the outcome of main symptoms and on validity parameters in screening patients suitable for shunting. Methods: In February 2015 we searched electronic databases from their inception to the current date, using the following key words: normal pressure hydrocephalus, idiopathic normotensive hydrocephalus, shunt operation, CSF tap test, predictive value, validity. The search retrieved 8 articles explicitly addressing the topic. Results: There was a very high positive predictive value of CSF-TT: 92% (range from 73% to 100%) but a low negative predictive value: 37% (18%–50%). Also, the CSF-TT has high specificity: 75% (33%–100%) but average sensitivity: 58% (26%–87%). The overall accuracy of the test was 62% (45%–83%). Conclusions: This systematic review did not provide unambiguous validity of the CSF-TT in the screening of patients for shunting. The validity of the CSF-TT is good for patient inclusion for shunting due to the fact that the positive response to the test is very reliable. Unfortunately, the negative response to the test does not reliably make these patients ineligible for shunting. Further studies are needed to improve and standardize the methodology in order to optimize the detection power of the test. © 2016 Elsevier B.V. All rights reserved.
Contents 1. 2.
3.
4.
Introduction . . . . . . . . . . . . . . . . Methods . . . . . . . . . . . . . . . . . . 2.1. Study design . . . . . . . . . . . . . 2.1.1. Search strategy. . . . . . . . 2.2. Study selection . . . . . . . . . . . . 2.2.1. Inclusion and exclusion criteria 2.2.2. Data extraction. . . . . . . . 2.3. Assessment of methodological quality . 2.3.1. Diagnostic accuracy measures . Results . . . . . . . . . . . . . . . . . . . 3.1. Study selection . . . . . . . . . . . . 3.1.1. Study characteristics . . . . . 3.2. Diagnostic accuracy measures . . . . . 3.2.1. Methodological quality . . . . 3.2.2. Characteristics of studies . . . Discussion . . . . . . . . . . . . . . . . . 4.1. Limitations. . . . . . . . . . . . . . 4.2. Suggestions for future research . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
⁎ Corresponding author at: Department of Neurology, University Hospital Split, Spinčićeva 1, 21000 Split, Croatia. E-mail address: [email protected] (M. Mihalj).
http://dx.doi.org/10.1016/j.jns.2016.01.028 0022-510X/© 2016 Elsevier B.V. All rights reserved.
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
79 79 79 79 79 79 80 80 81 81 81 81 81 81 81 82 83 83
M. Mihalj et al. / Journal of the Neurological Sciences 362 (2016) 78–84
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Declaration of interests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction In 1965, Hakim and Adams described the syndrome of gait disturbance, cognitive deterioration and urinary incontinence — a clinical triad associated with ventricular enlargement disproportional to any sulcal enlargement (distinguishing it from atrophy), in the absence of elevated cerebrospinal fluid (CSF) pressure during lumbar puncture [1,2]. Normal pressure hydrocephalus (NPH) in patients without known precipitants is termed primary or idiopathic iNPH (iNPH). When it occurs after other diseases, such as meningitis, traumatic brain injury, subarachnoid hemorrhage, cerebral infarction, this syndrome is called secondary NPH [3,4]. Some authors point to a hereditary predisposition towards iNPH [5,6]. There is variation in the clinical presentation, severity and progression of these symptoms, and it is not necessary for the entire triad to be present in order to consider diagnosing iNPH. Gait and balance impairment appear either before or concurrently with urinary incontinence or the onset of dementia. Symptoms are developed insidiously, and generally occur between the sixth and eighth decade of life [7,8]. Gait disturbances are the first signs of iNPH, and described as apraxic, glue-footed, magnetic, bradykinetic, and shuffling gait [9,10,11]. They are often misinterpreted as symptoms of Parkinson's disease [12]. Urinary incontinence usually follows gait abnormalities and almost always includes urinary urgency [13,14]. Dementia is rarely the first and foremost symptom of NPH, although it is often present [9, 13]. The impairment is mainly cognitive and subcortical in type, characterized by inattention, delay in responding and remembering, lack of spontaneity but without cognitive decline as in cortical dementia [13, 15]. Although NPH is commonly referred to as a treatable form of dementia, cognitive deficits and memory loss are the symptoms less likely to respond to shunting [16,17,18]. Neuroimaging with CT or MRI is an essential part of the evaluation of patients with suspected NPH/iNPH and ventricular enlargement is necessary to establish the diagnosis of NPH for patients with appropriate symptoms. A frontal horn ratio (Evans' index), defined as the maximal frontal horn ventricular width divided by the transverse inner diameter of the skull, indicates ventriculomegaly if it is 0.3 or greater [10,18]. Diagnosis based on clinical and radiological signs alone can be problematic and additional diagnostic testing may be required to determine which patients could benefit from shunting. A transient, favorable clinical effect in three NPH patients after removal of 15 ml of fluid was first described by Adams et al. [2]. CSF-TT was later modified by Wikkelso et al. by introducing the removal of larger quantities of fluid (40–50 ml) and quantitative testing of the main symptoms [19,20]. Although inexpensive, readily available and safe (without serious complications), there are controversial opinions about CSF-TT's detection power and validity in screening patients suitable for shunting. We searched the Cochrane Database of Systematic Reviews (CDSR), but no systematic review on this topic was found. Therefore, a systematic review was conducted to determine clinical impact and validity of CSF-TT in the prediction of response to shunting. 2. Methods 2.1. Study design The methods of this systematic review were decided a priori and adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [21]. The PRISMA statement
79
83 83 84
includes a 27-item checklist designed to improve reporting of systematic reviews and meta-analyses. 2.1.1. Search strategy Several online databases (PubMed, OvidMedline, Current Contents, Science Direct, Ebsco, and Web of Science) were searched from their inception to February 30, 2015. Authors (M.M., K.D., K.K.,V.L.) typed keywords independently from one another using the Boolean operator OR & AND and searched the online databases. The search strategy was as follows: ((normal pressure hydrocephalus) OR (normotensive hydrocephalus) OR (idiopathic normal pressure hydrocephalus)) AND (shunt operation) AND ((CSF tap test) OR (spinal tap test)) AND predictive value. A total of 2898 published articles were searched by abstract or full text. Retrospective and prospective studies were also searched. There were no language restrictions. Any published article that explicitly addressed CSF-TT was included, particularly if the article provided numerical data that could be included in a systematic review. A meta-analysis could not be performed because most of the articles were not designed as a randomized controlled trial. The search strategy is shown in Fig. 1. A complementary manual search of reference lists and personal resources was also performed to identify any relevant articles missed in the electronic searches. Finally, we searched for associated publications of retrieved articles to obtain the most complete and up-to-date study results. Papers that did not present clinical data were excluded, and any duplicate presentation of clinical data was identified. 2.2. Study selection Two authors (M.M. and K.D.) reviewed the electronic database search results (title and abstract) independently. Any titles and abstracts that appeared to meet inclusion criteria were selected for full text review. The reference lists of the identified studies were reviewed to discover additional potentially eligible studies. Unpublished data and conference proceedings were excluded from this review. Abstracts were excluded when both investigators agreed they were not relevant. Any disagreements were resolved within discussion between the two authors. The same two authors independently conducted the full text review of the retrieved articles in regard to the inclusion and exclusion criteria. Any disagreements were resolved by discussion, with a third and fourth author (K.K., V.L.) consulted if resolution was not achieved, to produce the final articles for inclusion. A data extraction form was prepared, with two authors (M.M., K.D.) independently extracting data from the selected studies. Authors (M.M., K.K.) reviewed the completed form for accuracy, with any disagreements resolved by the fourth author (V.L.). 2.2.1. Inclusion and exclusion criteria In our analysis, we included studies that met the following prespecified criteria: a diagnosis of NPH (iNPH) based on clinical examination and neuroimaging (CT/MRI), outcome data for at least for two of the three main symptoms (gait, cognition, continence) after CSF-TT and shunting, follow up and reevaluation of outcomes (at least once and not earlier than three months after shunting). NPH (iNPH) was defined as a symmetrical quadriventricular enlargement (Evans' Index ≥ 0.3) without clinically significant cortical and parenchymal lesions (atrophy, infarcts) with free communication between the ventricular system and the subarachnoid space (“communicating”
80
M. Mihalj et al. / Journal of the Neurological Sciences 362 (2016) 78–84
Fig. 1. Flow diagram for identification of relevant studies.
hydrocephalus) verified by CT/MR neuroimaging. Also, for the diagnosis to be made we required the presence of at least two symptoms of the clinical triad (gait was mandatory) Studies were not included if they did not provide data on the main symptoms and parameters of validity of the CSF-TT or raw data on outcomes of the main symptoms after CSF-TT and shunting from which it was possible to assess the parameters of validity. Furthermore, studies which did not describe (quantitatively and qualitatively) the reasons why patients were lost to follow-up were not included. Also, studies where shunting outcomes were not taken as reference (“gold standard”) for the definitive diagnosis and treatment of NPH were excluded.
symptoms), inclusion and exclusion criteria, the setting for patient recruitment - whether recruitment was consecutive and/or data collection was performed prospectively, data on the diagnosis of NPH (clinical, neuroimaging — CT/MRI), data on the outcomes for the main symptoms (gait, cognition, continence), number of patients lost to follow-up. We also collected: description and timing of outcome measures, data on the amount of removed CSF, CSF opening pressure, and complications after CSF-TT.
2.2.2. Data extraction The following data were extracted: author, year of publication, characteristics of the study population (age, gender, prevalence of the main
The methodological quality of the studies was assessed by the 2 reviewers (M.M., K.D.) using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool [22].
2.3. Assessment of methodological quality
M. Mihalj et al. / Journal of the Neurological Sciences 362 (2016) 78–84
The reviewers were blind to each other's assessment and scored the criteria items with “yes” or “no” when studies satisfied or failed to meet the criteria, respectively, and “unclear” when information was lacking to decide whether the study satisfied or met that specific item. In the case of disagreement, the 2 reviewers tried to reach a consensus on each criterion, and in case of persisting disagreement, a third or fourth reviewer (K.K. and V.L.) was involved. 2.3.1. Diagnostic accuracy measures For quantitative assessment, statistical measures of predictive validity (sensitivity, specificity, positive and negative predictive values, accuracy) from each included study were extracted or calculated from cumulative (2 × 2) tables completed by one of the authors (M.M.) and confirmed by another author (K.D.). 3. Results 3.1. Study selection Fig. 1 depicts the flow of articles through the review process. Eight articles, evaluating a total of 482 patients, met the inclusion criteria for this review. Seven of these articles were identified through the electronic database search [20,24–29] while the remaining article [23] was identified by reference searching. The results were heterogeneous and varied from very good [20,25, 28] to disappointingly limited power of the CSF-TT [24,26,29] in predicting the response to shunting. 3.1.1. Study characteristics Tables 1 and 2 present the characteristics of the studies included in the review. 3.2. Diagnostic accuracy measures Table 3 presents the parameters of the CSF-TT's validity in the included studies. 3.2.1. Methodological quality Fig. 2 presents the QUADAS 2 assessment results for each included study. Five studies were found to have low risk of bias [20,25,26,28, 29], and three studies were deemed to be at risk of bias [23,24,27]. Six studies were deemed to have low concern regarding applicability [20,
81
25,26,27,28,29], whereas two studies had concerns regarding applicability [23,24].
3.2.2. Characteristics of studies Haan and Thomeer examined the effect of the CSF-TT and external lumbar drainage (ELD) on main NPH symptoms in 32 iNPH patients. The CSF-TT had a low sensitivity (43%) and negative predictive value (42%) but specificity and positive predictive value were very high, 100% [23]. Wikkelsø and colleagues [20] studied the predictive value in 27 patients (21 NPH/6 iNPH), 15 patients were included from the previous study by the same authors and reported good sensitivity (68%), lower negative predictive value (45%) but positive predictive value and specificity was very high-100%. Malm et al. [24] compared the results after CSF-TT and CSF flow conductance with the outcomes after shunt surgery in 35 iNPH cases and showed low negative predictive value (23%) and specificity (33%) while the positive predictive value was 73% and sensitivity was 62%. Damasceno et al. [25] reported very high values for positive predictive value (93%) and sensitivity (87%), specificity was 67% but the negative predictive value was 50%. Kahlon et al. [27] compared results between the lumbar infusion test (LIT) and CSF-TT with outcomes after shunting in 47 (35 iNPH/12 NPH) patients. The total number of false negative values was unknown because NPH patients, with negative results in both LIT and CSF-TT (21/47), were not operated on. Therefore, the exact values for sensitivity and specificity could not be calculated. For those who were operated on, parameters of validity were as follows: positive predictive value — 94%, specificity — 88.9%, negative predictive value — 26.7% and sensitivity — 42%. Walchenbach et al. studied the effect of external lumbar drainage (ELD) and CSF-TT in 47 patients and showed that CSF- TT had low sensitivity (26%) and negative predictive value (32%). On the contrary, it had very high specificity and positive predictive value - 100% [26]. Ishikawa and colleagues analyzed the results of the effect of the tap test from the SINPHONI (Study of Idiopathic Normal Pressure Hydrocephalus on Neurological Improvement). The main conclusion was that the sensitivity and specificity of the CSF-TT were optimum when improvement on any iNPH grading scale was combined with CSF opening pressure ≥ 15 cm H20 [28]. Therefore sensitivity rose from 71% to 83% while specificity remained at 65%. The positive and negative predictive values were 89% and 50%, respectively [28]. In the Eu-iNPH study by Wikkelsø et al. a very low negative predictive value (18%) was obtained due to the large number of false negatives. The test sensitivity was 52%. The
Table 1 Demographic data, diagnostic methods, follow - up time and prevalence of main NPH symptoms. Study
Age (years) Mean ± (SD) or range
Gender (F/M)
Number of patients/lost iNPH (NPH) diagnose to follow-upa
Follow-up after shunt (months)
Cognition Gait disturbance impairment (%) (%)
Incontinence (%)
Wikkelsø et al. [29] Ishikawa et al. [28] Walchenbach et al. [26] Kahlon et al. [27] Damasceno et al. [25] Malm et al. [24] Haan and Thomeer [23] Wikkelsø et al. [20]
71 (30–87)
69/73
142/27
Clinically & radiologically (CT/MR)
1, 3, 6, 12
Missing data
Missing data
75 (71–78)
42/58
100
Clinically & radiologically (MR)
3, 6, 12
Missing data 91%
80%
60%
77 (55–87)
14/37
48/1
Clinically & radiologically (MR)
2, 6, 12
Missing data
Missing data
72 ± 9
39/29
81/13
Clinically & radiologically (CT)
6 (4,5)
Missing data 91%
76%
60%
18
Clinically & radiologically (CT)
3, 6,12
100%
83%
83%
a
66.5 ± 9.6
6/12
71.7 ± 5.2
10/27
35/2
Clinically & radiologically (CT)
3
100%
49%
59%
8/24
32/6
Clinically & radiologically (CT)
3, 6, 12, 24
100%
90%
78%
14/13
27/3
Clinically & radiologically (CT/radionuclide cisternography)
3–6
89%
74%
81%
70.5 (58–80)
63 (40–81)
Number of patients who underwent CSF-TT/lost to follow-up.
82
M. Mihalj et al. / Journal of the Neurological Sciences 362 (2016) 78–84
Table 2 Data on the amount of CSF removed, opening pressure, outcome measures, time of evaluation and possible complications. Author
Amount of CSF (ml)
Opening CSF pressurea
Gait (Outcome measuresb)
Cognition (Outcome measuresb)
Continence (Outcome measuresb)
Complications
Wikkelsø et al. [20]
50
18 m walking test (3×), 2 h after CSF TT
Identical forms test, Bingleys memory test, reaction time test, 2 h after CSF TT
Missing data
No complication
Malm et al. [24] Damasceno et al. [25] Kahlon et al. [27] Walchenbach et al. [26] Ischikawa et al. [28] Wikkelsø et al. [29]
Different
16.32 ± 6.8 (5.44–29.4) cm H20 0 cm H20
2 headaches
b18 cm H20
50
Unknown
40
Unknown
30
12 (9–14) cm H20 and ≥15 13.6 (2.7–24.5)
MMSE, series of neuropsychological tests (Bingley's Memory, Digit Span, Peg Bord) Test of visuo-motor speed, visuo-constructive skills, memory test, 2–8 h after CSF TT Reaction time test, identical forms test, memory test-Bingley's charts, 1–3 h after CSF TT MMSE, series of neuropsychological tests, 6–8 h after CSF TT iNPHGS (0–4), MMSE, 7 days after CSF-TT
Missing data
50
25 m walking test (3×), Barthel index of ADL, 3 to 4 h after CSFTT 18 m walking test (4×), 2–8 h after CSF TT 18 m walking test (3×), 1–3 h after CSF-TT An ordinal (0–5) rating scale, 6–8 h after CSF TT iNPHGS(0–4), mRS, TUG, 1–2 days after CSF TT mRS, iNPH rating scale (1–8), 3 h after CSF TT
a b
50
cm H20
MMSE, three neuropsychological tests, 3 h after CSF TT
Missing data
No complication Missing data No complication An ordinal (3 point) No rating scale complication iNPHGS (0–4) No 7 days after CSF TT complication iNPH rating scale No (1–6), 3 h after CSF TT complication
Presented as mean value or range. Tests for evaluating outcomes before and after CSF-TT and shunting.
positive predictive value and specificity were 88% and 59%, respectively [29].
4. Discussion Our systematic review of eight prospective cohort studies, involving 482 cases of normal pressure hydrocephalus, demonstrated a questionable validity of the CSF-TT in predicting response to shunt surgery. This was particularly true in regard to the negative response to the test. The positive predictive value of CSF-TT was very high across the majority of analyzed studies, ranging between 73% [24] and 100% [20,23,26]. It was due to the fact that most patients - 80% (range 73%–86%) improved after shunting, which gave a high positive predictive value simply by chance. Similarly, due to the low percentage of false positives [20,23,26], specificity was also high (75%), ranging between 33% [24] and 100% [20,23,26]. This highlights the utility of CSF-TT as a rule-in test for shunting when positive. Only the study by Malm et al. showed a low specificity (33%) (many false positives), but also a low negative predictive value — 23% (many false negatives) [24]. The main reason for this finding was the removal of different amounts of CSF in different patients depending on the volume and elasticity of the system [24]. In contrast, the sensitivity was extremely heterogeneous, ranging between 26% in the study by Walchenbach et al. [26] and 87% in the study by Damasceno et al. [25]. Also, the negative predictive value was low, ranging between 18% in the study by Wikkelsø et al. [29] and 50% in the studies by Damasceno and Ishikawa et al. [25,28]. The accuracy was 62%, ranging between 45% in the study by Walchenbach et al. [26] and 83% in the study by Damasceno et al. [25].
A key source of the heterogeneity of the results for these parameters of validity of the CSF-TT is the lack of standardization of methodology in assessing the main symptoms of NPH in the analyzed studies. The amount of CSF removed varied from 30 ml, in the study by Ishikawa et al. [28] and 40 ml in the study by Walchenbach et al. [26] to 50 ml in other studies [20,25,27,29]. In the study by Malm et al. [24] different amounts of CSF were removed, and the study of Haan and Thomeer did not state this data [23]. It seems that the amount of CSF removed affects certain parameters of the test validity; higher sensitivity and negative predictive values were observed in the studies where greater amounts of cerebrospinal fluid were removed [20,25] than in the studies of Malm [24], and Walchenbach [26]. This was also highlighted by Kilic et al. [30] and Damasceno [31]. They introduced the removal of larger amounts of fluid by repeating CSF-TT (RTT) during two or three consecutive days, removing 30– 50 ml/d, to the total of 100–120 ml. The authors reported improved sensitivity, negative and positive predictive value, actually empowering the detection of responders for shunting. Since the test has no serious complications compared with shunt surgery, some authors recommend repeating the test [32]. The positive response to RTT is recommended as one of the predictors of successful response to shunting, as recently published by the American Academy of Neurology (AAN) practice guideline for predictors of shunting effectiveness and utility of shunting in iNPH [33]. Conversely, a disappointingly low negative predictive value (18%) is reported in the Eu-iNPH study by Wikkelsø et al., despite the volume of the removed fluid. The authors concluded that the power of CSF-TT to predict a good outcome of shunting was disappointingly limited [29]. Also, the heterogeneity between analyzed studies exists in terms of
Table 3 Parameters of validity of CSF-TT in assessing suitable patients for shunting. Author
Region
Total number of patients, design of study
PPV (%)
NPV (%)
SN (%)
SP (%)
AC (%)
Wikkelsø et al. [20] Haan and Thomeer [23] Malm et al. [24] Damasceno et al. [25] Kahlon et al. [27] Walchenbach et al. [26] Ishikawa et al. [28] Wikkelsø et al. [29]
Europe, Sweden Europe, The Netherlands Europe, Sweden South America, Brazil Europe, Sweden Europe, The Netherlands Asia, Japan Europe, 9 countries
27, prospective 32, prospective 35, prospective 18, prospective 81, prospective 47, prospective, 3 centers 100, prospective, 26 centers 142, single blinded, prospective, 13 centers
100 100 73 93 94 100 89 88
45 42 23 50 Unknowna 32 50 18
68 43 62 87 Unknowna 26 71 52
100 100 33 67 Unknowna 100 65 59
75 54 54 83 – 45 70 53
Legend: PPV = positive predictive value; NPV = negative predictive value; SN = sensitivity, SP = specificity; AC = accuracy. a Unknown number of negatives (false and true).
M. Mihalj et al. / Journal of the Neurological Sciences 362 (2016) 78–84
83
Fig. 2. Graphical presentation for QUADAS-2 results in diagnostic accuracy.
the CSF opening pressure, ranging from 0 cm H2O in the study of Malm et al. [24] to 29.4 cm H2O in the studies by Wikkelsø et al. [20]. In the two studies, data were not presented on the CSF opening pressure [26,27]. In the studies by Damasceno [25] and Ishikawa [28], where the opening pressures were set higher, higher negative predictive values (50%) and sensitivity were obtained (87% and 83%, respectively). Although the higher CSF opening pressure could indicate the possibility of secondary forms of NPH, Ishikawa and colleagues [28] pointed to the fact that there was a drop in false negatives when the CSF opening pressure was set at ≥ 15 cm H2O. Therefore, the sensitivity increased from 71% to 83%, while the specificity remained the same (65%). In clinical practice, fewer shunt responders would be missed. There was tremendous heterogeneity in the labeling of “improved” gait among analyzed studies. In the study by Haan and Thomeer, the gait improvement was too strictly categorized as mobile vs. immobile [23], while in the study by Walchenbach et al. [26] the scoring was more lenient (e.g., improvement for at least one point on the fivepoint gait score was considered markedly positive). Depending on the stipulated cut-off level for „significant improvement“, different parameters were obtained for the validity of CSF-TT among analyzed studies. For example, applying more stringent criteria or limits for the labeling of “improved” gait increases the number of false negatives and consequently lowers the negative predictive value and sensitivity, but also the accuracy of the test, as highlighted in recent studies [28,29]. Only in two recent studies, the newly developed iNPH scale was used, reducing the discrepancy in the assessment of outcomes, especially in the gait domain [28,29]. It is known that NPH patients have daily fluctuations in the performance of cognitive tasks and movements [8,9,10]. The assessment time of the outcomes of the main symptoms, particularly in the gait and neuropsychological domains, was significantly different among the analyzed studies: 1–4h [20,24,27,29]; 2–8 h [25,26]; and 1–2 days, [28] after performing CSF-TT. Only in the study by Ishikawa et al. [28] neuropsychological and continence tests were performed seven days after the tap test. According to previously published studies, the fastest and most obvious improvements were observed in the gait domain after CSF-TT and shunting, but the severity and duration of the gait disturbance (and other symptoms) were negatively associated with favorable outcomes after CSFTT and shunting [23–25,29,33,34]. Virhammar et al. highlighted that better results in the gait domain were obtained when the assessment was carried out several hours after CSF-TT, repeatedly during 24 h. It is particularly important to adequately prevent or treat post-lumbar puncture pain, which has a negative effect on outcomes [35]. In most studies, the Mini Mental State Examination (MMSE) was used for the assessment of outcomes, with several other similar neuropsychological tests. MMSE is recommended in the early assessment, because it shows more prominent changes within the first day after CSF-
TT than other neuropsychological tests [36]. Unfortunately, only two studies recorded improvement in the neuropsychological domain after CSF-TT and positive correlation with improvement after shunting [20, 27]. There was no recovery in the continence domain after CSF-TT in any of the analyzed studies, [20,23–29], which was expected due to the more difficult and long-term recovery of this symptom. Also, there were no serious complications during and after the test. 4.1. Limitations The lack of a standardized methodology was one of the main challenges of this review, as evidenced by clinical and statistical heterogeneity. The general quality of the included articles was moderate. Two studies were considered to be at high risk of bias, one within the index test domain because of its interpretation and another study was within the flow and timing domain because the same reference standard was not used for all patients and not all samples included in the study had been accounted for in the analysis (the number of negatives, especially false negatives was unknown). One study was judged as “unclear” in the index test domain because of insufficient data. In two studies the applicability was considered to be of high concern with regard to execution and interpretation of the index test. Also, one of the shortcomings of the analyzed studies was the lack of data on the extent and severity of the impairment for each of the main symptoms at baseline, which certainly affected the results of the outcomes. All of the above may have resulted in misclassification and estimated parameters of validity may have been biased. 4.2. Suggestions for future research In future, study methodology, including the measuring of the amount of CSF removed, opening pressure, cut-off level for labeling “improved” and timing of outcome assessment of the main symptoms after the tap test, should be standardized in order to reduce the risk of bias and enable better comparability. 5. Conclusions Inexpensive, readily available, and without serious side-effects, CSFTT is still the first choice of additional tests in the detection of suitable patients for shunt surgery. Due to lower reliability of the negative findings (low negative predictive value and sensitivity), the test should not be used for excluding patients from shunt surgery because many shunt responders could be missed. Declaration of interests The authors declare that they have no conflict of interest.
84
M. Mihalj et al. / Journal of the Neurological Sciences 362 (2016) 78–84
References [1] S. Hakim, R.D. Adams, The special clinical problem of symptomatic hydrocephalus with normal cerebrospinal fluid pressure: observations on cerebrospinal fluid hydrodynamics, J. Neurol. Sci. 2 (1965) 307–327. [2] R.D. Adams, C.M. Fisher, S. Hakim, R.G. Ojemann, W.H. Sweet, Symptomatic occult hydrocephalus with “normal” cerebrospinal-fluid pressure: a treatable syndrome, N. Engl. J. Med. 273 (1965) 117–126. [3] A.O. Hebb, M.D. Cusimano, Idiopathic normal pressure hydrocephalus: a systematic review of diagnosis and outcome, Neurosurgery 49 (2001) 1166–1184. [4] B. Beyerl, P.M. Black, Posttraumatic hydrocephalus, Neurosurgery 15 (1984) 257–261. [5] M.D. Cusimano, D. Rewilak, D.T. Stuss, J.C. Barrera-Martinez, F. Salehi, M. Freedman, Normal-pressure hydrocephalus: is there a genetic predisposition? Can J. Neurol. Sci. 38 (2011) 274–281. [6] A. McGirr, M.D. Cusimano, Familial aggregation of idiopathic normal pressure hydrocephalus: novel familial case and a family study of the NPH triad in an iNPH patient cohort, Neurol. Sci. 321 (2012) 82–88. [7] C.M. Fisher, The clinical picture in occult hydrocephalus, Clin. Neurosurg. 24 (1977) 270–284. [8] M.J. McGirt, G. Woodworth, A.L. Coon, G. Thomas, M.A. Williams, D. Rigamonti, Diagnosis, treatment, and analysis of long-term outcomes in idiopathic normal-pressure hydrocephalus, Neurosurgery 57 (2005) 699–705. [9] P.L. Gleason, P.M. Black, M. Matsumae, The neurobiology of normal pressure hydrocephalus, Neurosurg. Clin. North Am. 4 (1993) 667–675. [10] N. Relkin, A. Marmarou, P. Klinge, M. Bergsneider, P.M. Black, Diagnosing idiopathic normalpressure hydrocephalus, Neurosurgery 57 (2005) S4–S16. [11] C.M. Fisher, Hydrocephalus as a cause of disturbances of gait in the elderly, Neurology 32 (1982) 1358–1363. [12] M. Mihalj, M. Titlić, A. Marović, S. Srdelić-Mihalj, Gait apraxia, Bratisl. Lek. Listy 111 (2010) 101–102. [13] J.A. Vanneste, Diagnosis and management of normalpressure hydrocephalus, J. Neurol. 247 (2000) 5–14. [14] A. Larsson, C. Wikkelsö, M. Bilting, H. Stephensen, Clinical parameters in 74 consecutive patients shunt operated for normal pressure hydrocephalus, Acta Neurol. Scand. 84 (1991) 475–482. [15] P. Bret, J. Guyotat, J. Chazal, Is normal pressure hydrocephalus a valid concept in 2002? A reappraisal in five questions and proposal for a new designation of the syndrome as “chronic hydrocephalus”, J. Neurol. Neurosurg. Psychiatry 73 (2002) 9–12. [16] J.K. Krauss, M. Faist, M. Schubert, J.J. Borremans, C.H. Lücking, W. Berger, Evaluation of gait in normal pressure hydrocephalus before and after shunting, Adv. Neurol. 87 (2001) 301–310. [17] P.M. Black, Idiopathic normal-pressure hydrocephalus. Results of shunting in 62 patients, J. Neurosurg. 52 (1980) 371–377. [18] R.C. Petersen, B. Mokri, E.R. Laws Jr., Surgical treatment of idiopathic hydrocephalus in elderly patients, Neurology 35 (1985) 307–311. [19] C. Wikkelsø, H. Andersson, C. Blomstrand, G. Lindqvist, The clinical effect of lumbar puncture in normal pressure hydrocephalus, J. Neurol. Neurosurg. Psychiatry 45 (1982) 64–69.
[20] C. Wikkelsø, H. Andersson, C. Blomstrand, G. Lindqvist, P. Svendsen, Normal pressure hydrocephalus: predictive value of the cerebrospinal fluid tap-test, Acta Neurol. Scand. 73 (1986) 566–573. [21] D. Moher, A. Liberati, J. Tetzlaff, D. Altman, PRISMA GROUP, Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement, PLoS Med. 6 (2009). http://dx.doi.org/10.1371/journal.pmed.1000097. [22] P. Whiting, A. Rutjes, M. Westwood, S. Mallett, J. Deeks, J. Reitsma, et al., QUADAS-2 Group, QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies, Ann. Intern. Med. 155 (2011) 529–536. [23] J. Haan, R.T. Thomeer, Predictive value of temporary external lumbar drainage in normal pressure hydrocephalus, Neurosurgery 22 (1988) 388–391. [24] M. Malm, B. Kristensen, T. Karlsson, M. Fagerlund, J. Elfverson, J. Ekstedt, The predictive value of cerebrospinal fluid dynamic tests in patients with the idiopathic adult hydrocephalus syndrome, Arch. Neurol. 52 (1995) 783–789. [25] B.P. Damasceno, E.F. Carelli, D.C. Honorato, J.J. Facure, The predictive value of cerebrospinal fluid tap-test in normal pressure hydrocephalus, Arq. Neuropsiquiatr. 55 (1997) 179–185. [26] R. Walchenbach, E. Geiger, R.T. Thomeer, J.A. Vanneste, The value of temporary external lumbar CSF drainage in predicting the outcome of shunting on normal pressure hydrocephalus, J. Neurol. Neurosurg. Psychiatry 72 (2002) 503–506. [27] B. Kahlon, G. Sundbärg, S. Rehncrona, Comparison between the lumbar infusion and CSF tap tests to predict outcome after shunt surgery in suspected normal pressure hydrocephalus, J. Neurol. Neurosurg. Psychiatry 73 (2002) 721–726. [28] M. Ishikawa, M. Hashimoto, E. Mori, N. Kuwana, H. Kazui, The value of the cerebrospinal fluid tap test for predicting shunt effectiveness in idiopathic normal pressure hydrocephalus, Fluids Barriers CNS 13:9 (1) (2012) 1. [29] C. Wikkelsø, P. Hellström, P.M. Klinge, J.T. Tans, European iNPH Multicentre Study Group, 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 (2013) 562–568. [30] K. Kilic, A. Czorny, J. Auque, Z. Berkman, Predicting the outcome of shunt in normal pressure hydrocephalus, J. Clin. Neurosci. 14 (2007) 729–736. [31] B.P. Damasceno, Normal pressure hydrocephalus. Diagnostic and predictive evaluation, Dement. Neuropsychol. 3 (2009) 8–15. [32] T.S. Lim, S.W. Yong, S.Y. Moon, Repetitive lumbar punctures as treatment for normal pressure hydrocephalus, Eur. Neurol. 62 (2009) 293–297. [33] J.J. Halperin, R. Kurlan, J.M. Schwalb, M.D. Cusimano, G. Gronseth, D. Gloss, Practice guideline: idiopathic normal pressure hydrocephalus: response to shunting and predictors of response: report of the guideline development, dissemination, and implementation subcommittee of the American academy of neurology, Neurology 85 (2015) 2063–2071. [34] U. Meier, J. Lemcke, U. Neumann, Predictors of outcome in patients with normalpressure hydrocephalus, Acta Neurochir. 96 (2006) 352–357. [35] J. Virhammar, K.G. Cesarini, K. Laurell, The CSF tap test in normal pressure hydrocephalus: evaluation time, reliability and the influence of pain, Eur. J. Neurol. 19 (2012) 271–276. [36] H. Schmidt, J. Elster, I. Eckert, J. Wiefek, W. Paulus, N. von Steinbuechel, et al., Cognitive functions after spinal tap in patients with normal pressure hydrocephalus, J. Neurol. 261 (2014) 2344–2350.