- Email: [email protected]
Validation of a Predictive Scoring System for Ventriculoperitoneal Shunt Insertion After Aneurysmal Subarachnoid Hemorrhage
Accepted Manuscript Validation of a Predictive Scoring System for Ventriculoperitoneal Shunt Insertion following Aneurysmal Subarachnoid Hemorrhage Raghav Gupta, B.S., Luis C. Ascanio, M.D., Alejandro Enriquez-Marulanda, M.D., Christoph J. Griessenauer, M.D., Anu Chinnadurai, Ray Jhun, B.S., Abdulrahman Alturki, M.B.B.S., M.Sc., FRCSC, Christopher S. Ogilvy, M.D., Ajith J. Thomas, M.D., Justin M. Moore, M.D., Ph.D. PII:
S1878-8750(17)31656-X
DOI:
10.1016/j.wneu.2017.09.140
Reference:
WNEU 6584
To appear in:
World Neurosurgery
Received Date: 22 August 2017 Revised Date:
19 September 2017
Accepted Date: 20 September 2017
Please cite this article as: Gupta R, Ascanio LC, Enriquez-Marulanda A, Griessenauer CJ, Chinnadurai A, Jhun R, Alturki A, Ogilvy CS, Thomas AJ, Moore JM, Validation of a Predictive Scoring System for Ventriculoperitoneal Shunt Insertion following Aneurysmal Subarachnoid Hemorrhage, World Neurosurgery (2017), doi: 10.1016/j.wneu.2017.09.140. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1 Validation of a Predictive Scoring System for Ventriculoperitoneal Shunt Insertion
2
following Aneurysmal Subarachnoid Hemorrhage
3
Raghav Gupta, B.S., * Luis C. Ascanio, M.D., * Alejandro Enriquez-Marulanda, M.D., Christoph
4
J. Griessenauer, M.D., Anu Chinnadurai, Ray Jhun, B.S., Abdulrahman Alturki, M.B.B.S.,
5
M.Sc., FRCSC, Christopher S. Ogilvy, M.D., Ajith J. Thomas, M.D., Justin M. Moore, M.D.,
6
Ph.D.
7
Neurosurgical Service, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston,
8
Massachusetts, USA
9
*These authors contributed equally to this study.
M AN U
SC
RI PT
1
10
Abbreviations: Attached on a separate sheet.
12
Keywords: chronic hydrocephalus; venticuloperitonal shunt; subarachnoid hemorrhage; scoring
13
system
14
TE D
11
Corresponding Author
16
Justin M. Moore, M.D., Ph.D.
17
Neurosurgical Service
18
Beth Israel Deaconess Medical Center
19
Harvard Medical School
20
110 Francis Street
21
Boston, MA, 02215
22
Email: [email protected]
23
Phone: 617-632-7246
24
AC C
EP
15
ACCEPTED MANUSCRIPT 2 25
Abstract
26 Introduction
28
Hydrocephalus is a frequently encountered complication in the context of aneurysmal
29
subarachnoid hemorrhage (aSAH). Here, we performed an external validation of the recently
30
proposed Post-Subarachnoid Shunt Scoring (PS3) system, which aims to stratify patients
31
presenting with aSAH based on their relative risk of requiring ventriculoperitoneal (VP) shunt
32
insertion.
SC
RI PT
27
M AN U
33 Materials and Methods
35
A retrospective review of all patients presenting with aSAH to our institution between July 2007
36
and December 2016, who underwent CT imaging at the time of hospital admission, was
37
performed.
38
TE D
34
Results
40
A total of 242 patients (66.1% female) with aSAH were included in the analysis with a mean age
41
of 55.6 years. Sixty-four (26.4%) patients had a Hunt and Hess grade of 4 or 5 on admission. An
42
external ventricular drain (EVD) was placed in 130 (53.7%) patients during the hospital
43
admission. EVD placement was found to correlate with an increased rate of VP shunt placement
44
(p<0.001), and a trend towards an association between a high Hunt and Hess grade and VP shunt
45
placement, was observed (p=0.05). The area under the ROC for the PS3 scoring system was
46
found to be 0.845. The system reliably predicted shunt-dependent chronic hydrocephalus in our
47
patient cohort (Odds ratio [OR]: 3.36; 95% confidence interval [CI]: 2.31-4.89; p <0.001).
AC C
EP
39
ACCEPTED MANUSCRIPT 3 48 Conclusion
50
Data from this study validated the previously proposed PS3 system, which was found to more
51
accurately predict shunt-dependent chronic hydrocephalus in patients with aSAH, as compared to
52
other such systems in the neurosurgical literature such as the CHESS, BNI, and SDASH systems.
RI PT
49
53 Word Count: 249
SC
54 55
AC C
EP
TE D
M AN U
56
ACCEPTED MANUSCRIPT 4 57
Introduction
58 Nearly 80% of nontraumatic subarachnoid hemorrhage (SAH) cases are attributable to ruptured
60
intracranial aneurysms.1 The resulting case fatality rate is greater than 50%, and patients who
61
survive often suffer from cognitive deficits and a subsequent deterioration in quality-of-life.2,3
62
Hydrocephalus is a frequently encountered complication in the context of SAH, and can occur in
63
up to 30% of patients.4 It is classified as being either acute (0-3 days post-SAH), subacute (4-13
64
days post-SAH), or chronic (≥14 days post-SAH).5 The incidence of chronic hydrocephalus after
65
SAH, which is typically managed through diversion of cerebrospinal fluid (CSF) flow via
66
ventriculoperitoneal (VP) shunt insertion, is reported to be between 6.0 and 48.0%.6,7
M AN U
SC
RI PT
59
67
Prior studies evaluating factors associated with VP shunt placement have found that a larger third
69
ventricular diameter at admission, history of sympathomimetic drug use, hyponatremia, high
70
Fisher computed tomography (CT) grade, low initial Glasgow Coma Scale (GCS) score, a higher
71
Hunt and Hess grade at admission, female gender, aneurysm location within the posterior
72
circulation, aneurysm size, and age, each predict the need for subsequent shunt insertion.6,8
73
However, in a recent study published by our group., a simpler scoring system based on external
74
ventricular drain (EVD) insertion, a high (≥4) Hunt and Hess grade at admission, and a modified
75
Fisher CT grade of 4 was proposed based on patients enrolled in the Cerebral Aneurysm Renin
76
Angiotensin System (CARAS) study.9 It was found to accurately predict the need for shunt
77
placement in a cohort of 149 patients presenting with aneurysmal SAH (aSAH). In the present
78
study, we aimed to validate this PS3 (Post-Subarachnoid Shunt Scoring) system using data from
AC C
EP
TE D
68
ACCEPTED MANUSCRIPT 5 79
a separate cohort of patients who presented with aSAH to a major academic center in the U.S.,
80
over an 8-year period, who had not been a part of the original CARAS study.
AC C
EP
TE D
M AN U
SC
RI PT
81
ACCEPTED MANUSCRIPT 6 82
Materials and Methods
83 Patient Selection
85
Institutional Review Board approval was obtained prior to the commencement this study. A
86
retrospective review of all consecutive patients presenting with SAH, between July 2007 and
87
December 2016, to an academic neurosurgical department in the U.S., was performed. Patients
88
were included if the etiology of their SAH was aneurysmal, and they had received a CT scan
89
upon being admitted. Patients were excluded if they presented with either non-traumatic non-
90
aneurysmal SAH, expired prior to initial hospital discharge, or if they had not undergone CT
91
imaging at the time of hospital admission (Figure 1).
M AN U
SC
RI PT
84
92
The variables collected for each patient included: patient demographics (age, gender, smoking
94
status, history of hypertension, history of antiplatelet/anticoagulant use), initial clinical
95
presentation (Hunt and Hess grade on admission prior to EVD insertion, Glasgow Coma Scale
96
(GCS) score on admission, presence of focal neurological deficits on admission, and Modified
97
Fisher CT scale grade on admission), aneurysm characteristics (aneurysm location, morphology,
98
presence of a daughter sac), EVD placement within the same hospital admission, treatment
99
modality, and modified Rankin Scale (mRS) score at discharge. The Hunt and Hess scale grade,
100
modified Fisher CT scale grade, and history of EVD insertion, were used to tabulate a score for
101
each patient as per the PS3 system.9
103
EP
AC C
102
TE D
93
External Ventricular Drain and Ventriculoperitoneal Shunt Placement Protocol
ACCEPTED MANUSCRIPT 7 The protocol at our institution is to place an EVD in all patients presenting with acute
105
hydrocephalus (confirmed via radiographic and/or clinical examination) or in patients with
106
suspected increased intracranial pressure (ICP) based on physical examination. The decision to
107
insert a VP shunt rather than to perform additional EVD clamp trials was based on multiple
108
criteria including the number of clamp trials that had been already performed, the severity of the
109
hydrocephalus as determined via CT imaging, and the rise in ICP once the EVD had been
110
clamped. Other factors including a high Hunt and Hess grade and modified Fisher CT grade at
111
admission, were used to guide decision-making. The clinical gestalt of the two cerebrovascular
112
neurosurgeons involved in the management of these patients, however, often heavily influenced
113
the decision to insert a shunt or to perform additional clamp trials.
M AN U
SC
RI PT
104
114 Statistical Analyses
116
The Chi-square test was used for the analysis of categorical variables. The t-test was used for
117
analyzing normally distributed continuous variables. A mean and standard deviation (SD) were
118
reported for normally distributed variables. The scoring system was validated via development of
119
a receiver operating characteristic (ROC) curve. The area under the curve was then calculated.
120
The Hosmer-Lemeshow test was performed to assess goodness of fit of the univariable logistic
121
regression model that was developed; goodness of fit was defined by a p-value >0.05. Statistics
122
were performed using the STATA software (StataCorp, Texas, U.S.). Significance was defined
123
with a p-value <0.05.
125
EP
AC C
124
TE D
115
ACCEPTED MANUSCRIPT 8 126
Results
127 Patients Characteristics
129
A total of 242 patients presented with aSAH to our institution and underwent CT imaging upon
130
hospital admission, within this time period. The mean age of these patients was 55.6 years
131
(Standard deviation[SD] ± 14.0 years). A history of hypertension or smoking was noted in 115
132
(50.2%) and 84 (42.9%) patients, respectively (Table 1). A majority (54.4%) of patients had
133
aneurysms located within the anterior circulation and of saccular morphology (92.4%)
SC
RI PT
128
M AN U
134 Clinical Presentation
136
Sixty-four (26.4%) patients had a Hunt and Hess grade of 4 or 5 on admission. A majority (149;
137
61.6%) of patients had a modified Fisher CT grade 4 hemorrhage. Thirty (12.6%) patients
138
presented with focal neurological deficits. Acute hydrocephalus was detected in 101 (41.9%)
139
patients. An EVD was placed in 130 (53.7%) patients. Fifty-four (22.3%) patients underwent VP
140
shunt placement for chronic hydrocephalus. Nine of these shunts were inserted in a separate
141
hospital admission due to the onset of delayed hydrocephalus while the remaining shunts were
142
inserted in the same admission period as the initial hemorrhage.
EP
AC C
143
TE D
135
144
Predictors of Ventriculoperitoneal Shunt Insertion
145
No differences in gender (p=0.57), age (p=0.47), history of hypertension (p=0.76), smoking
146
status (p=0.14), previous antiplatelet/anticoagulant use (p=0.29), and presence of neurologic
147
deficits at the time of admission (p=0.82), were noted in patients who did and did not undergo
148
VP shunt placement (Table 2). Aneurysm treatment modality was not found to be a predictor of
ACCEPTED MANUSCRIPT 9 VP shunt placement (p=0.43). Conversely, a higher Hunt and Hess grade at admission (p<0.001),
150
a lower GCS at admission (p<0.001), a higher modified Fisher CT grade hemorrhage (p<0.001),
151
EVD insertion during hospital admission (p<0.001), a higher mRS at discharge (p<0.001), and a
152
higher score on the PS3 system (p<0.001) were found to predict the need for VP shunt insertion.
153
Twenty-eight (51.8%) patients with a PS3 score of 4 underwent VP shunt placement as
154
compared to zero patients who had a PS3 score of 1 (Table 3). Patients with a PS3 score of 3 or 4
155
were significantly more likely to have a VP shunt inserted than patients with a score of 0, 1, or 2
156
(OR: 17.66; p<0.001).
157
M AN U
SC
RI PT
149
Multivariable Logistic Regression, Univariable Logistic Regression, and Receiver Operating
159
Characteristic Curve Analyses
160
On multivariable regression analysis, EVD placement (p<0.001) was found to correlate with an
161
increased rate of VP shunt placement (Table 4). A trend towards an association between high
162
Hunt and Hess grade at admission and VP shunt placement was observed (p=0.05). The area
163
under the ROC for the PS3 scoring system was found to be 0.845 (Figure 2). On univariable
164
logistic regression analysis, the PS3 score reliability predicted shunt-dependent chronic
165
hydrocephalus in our patient cohort (Odds ratio [OR]: 3.36; 95% confidence interval [CI]: 2.31-
166
4.89; p <0.001). The Hosmer-Lemeshow test demonstrated goodness of fit of the PS3 scoring
167
system (p=0.11).
EP
AC C
168
TE D
158
ACCEPTED MANUSCRIPT 10 169
Discussion
170 Background
172
Chronic hydrocephalus is a clinically important complication of aSAH, with up to 48% of
173
patients in whom EVDs are placed requiring VP shunt insertion for management of this
174
condition.6,7 Several studies in the neurosurgical literature identifying factors predictive of the
175
need for shunt insertion following aSAH have been published, albeit with significant
176
heterogeneity in regards to the individual factors that they have identified.5,6,8-14 A recent
177
systematic review and meta-analysis of these studies by Wilson et al., identified higher Fisher
178
grade (3/4 vs 1/2), presentation with acute hydrocephalus, in-hospital complications (including
179
pneumonia, meningitis, vasospasm, and ischemic strokes), intraventricular hemorrhage, high
180
Hunt and Hess grade, older age, female gender, and aneurysm location within the posterior
181
circulation as risk factors for shunt insertion.10 Of these factors, presentation with acute
182
hydrocephalus (OR: 5.67; 95% CI: 3.96-8.12) and a higher Fisher grade on CT imaging (OR:
183
7.74; CI: 4.47-13.41) were the two strongest predictors of the need for VP shunt placement.9
SC
M AN U
TE D
EP
184
RI PT
171
Previous Predictive Algorithms for Ventriculoperitoneal Shunt Insertion
186
Development of a simple system which could stratify patients presenting with aSAH based on
187
the risk of requiring surgical shunt placement, is essential. For high-risk patients, such a system
188
could potentially guide neurosurgeons towards replacing ventricular catheters with permanent
189
shunts more rapidly, leading to a decreased incidence of nosocomial EVD infections, fewer
190
clamp trial failures, shortened hospitals stays, and an overall decrease in treatment costs.15,16
191
Conversely, for patients at low risk of requiring VP shunt placement, a more aggressive EVD
AC C
185
ACCEPTED MANUSCRIPT 11 192
weaning process could be utilized as shunt placement is associated with several complications
193
including infections, shunt mechanical dysfunction, and/or obstruction.17
194 However, few such scoring systems exist within the neurosurgical literature. These include the
196
failure risk index (FRI) score,8 the Chronic Hydrocephalus Ensuing from SAH Score (CHESS),16
197
and the Shunt Dependency in aSAH (SDASH) score (Table 5).15 The FRI, however, has not been
198
widely used in the clinical setting.15 The Barrow Neurological Institute (BNI) scale,18,19 which
199
was originally developed to predict the incidence of symptomatic vasospasm following aSAH as
200
a function of the thickness of the blood clot, was also predictive of shunt-dependent chronic
201
hydrocephalus (area under ROC: 0.649).15 In the present study, we found the PS3 score more
202
accurately (OR: 3.36; 95% CI: 2.31-4.89; p <0.001; area under ROC 0.845) predicted the need
203
for shunt placement in a large cohort of patients, than either the SDASH (OR: 2.50; 95% CI:
204
1.86-3.35; p<0.001; area under ROC: 0.785), the CHESS (OR: 1.53; 95% CI 1.33-1.78; p<0.001;
205
area under ROC: 0.769), or the BNI (OR: 2.02; 95% CI: 1.45-2.82; p<0.001; area under ROC:
206
0.649) scoring systems.15 EVD placement in the same hospital admission was found to
207
independently predict the need for shunt placement even after age, gender, GCS score,
208
presentation with acute hydrocephalus, and aneurysm location were controlled for in a
209
multivariable regression model. A trend towards an association between a higher Hunt and Hess
210
grade at admission and shunt placement was observed as well (p=0.05). Both parameters are
211
components of the PS3 scoring system.
213
SC
M AN U
TE D
EP
AC C
212
RI PT
195
Clinical Utility of the PS3 Scoring System
ACCEPTED MANUSCRIPT 12 Adoption of a scoring system within the clinical setting for the rapid stratification of patients is
215
predicated on the simplicity of the system. The PS3 scoring system contains three factors, two of
216
which (Hunt and Hess & Modified Fisher CT score) are routinely assessed in patients presenting
217
with aSAH, and have been each independently found to predict the need for shunt insertion.10,20
218
The third (EVD insertion) is a binary variable (yes or no) which can be easily and rapidly
219
recorded. The modified Fisher CT scale assesses whether there is an intraventricular hemorrhage
220
(IVH) component, and is a variable listed separately in the more complex CHESS scoring
221
system. As obstruction of cerebrospinal fluid (CSF) flow has been proposed as a possible basis
222
for the development of hydrocephalus,14 the presence of IVH could serve as the physiological
223
mechanism by which this occurs. Additionally, the modified Fisher scale score is a more detailed
224
variant of the Fisher scale score, which was found to be the strongest predictor of shunt-
225
dependent hydrocephalus in Wilson et al.’s meta-analysis.10
M AN U
SC
RI PT
214
TE D
226 Limitations
228
This study is limited by its retrospective design. However, the PS3 system was validated in the
229
current study using a large cohort of patients who presented with aSAH to a different institution
230
than the one where the system was originally proposed. The decision to place a VP shunt was
231
made on a per-patient basis by the attending physician, and different neurosurgeons may have
232
various thresholds for shunt insertion which are difficult to assess.
AC C
233
EP
227
ACCEPTED MANUSCRIPT 13 234
Conclusion
235 Chronic hydrocephalus is a clinically important condition with a significant proportion of
237
patients who present with aSAH requiring eventual surgical shunt insertion. Few scoring systems
238
exist in the neurosurgical literature which can rapidly stratify patients based on their risk of
239
requiring VP shunt placement. The PS3 scoring system was recently proposed to address this,
240
and is based on Hunt and Hess grade, modified Fisher CT score, and EVD insertion. In the
241
present study, we performed an external validation of the PS3 score, and found it to more
242
accurately predict shunt-dependent chronic hydrocephalus than previous such scoring systems
243
such as the CHESS, BNI, and SDASH systems.
M AN U
SC
RI PT
236
AC C
EP
TE D
244
ACCEPTED MANUSCRIPT 14 245
Acknowledgments
246 Conflict of Interest Statement
248
The authors have no personal, financial, or institutional interest with regards to the authorship
249
and/or publication of this manuscript.
RI PT
247
250 Funding Statement
252
This research did not receive any specific grant from funding agencies in the public, commercial,
253
or not-for-profit sectors.
M AN U
SC
251
254
AC C
EP
TE D
255
ACCEPTED MANUSCRIPT 15
257 258 259 260
Additional Author Affiliations 1. C.J.G. Department of Neurosurgery, Geisinger Medical Center, Danville, Pennsylvania, USA 2. A.A. Department of Neurosurgery, National Neuroscience Institute, King Fahad Medical
RI PT
256
City, Riyadh, Saudi Arabia
AC C
EP
TE D
M AN U
SC
261
ACCEPTED MANUSCRIPT 16 262
References
263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304
1.
7.
8.
9. 10.
11.
12.
13.
14.
15.
RI PT
SC
6.
M AN U
5.
TE D
4.
EP
3.
AC C
2.
Suarez JI, Tarr RW, Selman WR. Aneurysmal subarachnoid hemorrhage. N Engl J Med. 2006;354(4):387-396. Mayer SA, Kreiter KT, Copeland D, et al. Global and domain-specific cognitive impairment and outcome after subarachnoid hemorrhage. Neurology. 2002;59(11):17501758. Hackett ML, Anderson CS. Health outcomes 1 year after subarachnoid hemorrhage: An international population-based study. The Australian Cooperative Research on Subarachnoid Hemorrhage Study Group. Neurology. 2000;55(5):658-662. Germanwala AV, Huang J, Tamargo RJ. Hydrocephalus after aneurysmal subarachnoid hemorrhage. Neurosurg Clin N Am. 2010;21(2):263-270. Vale FL, Bradley EL, Fisher WS, 3rd. The relationship of subarachnoid hemorrhage and the need for postoperative shunting. J Neurosurg. 1997;86(3):462-466. O'Kelly CJ, Kulkarni AV, Austin PC, Urbach D, Wallace MC. Shunt-dependent hydrocephalus after aneurysmal subarachnoid hemorrhage: incidence, predictors, and revision rates. Clinical article. J Neurosurg. 2009;111(5):1029-1035. Connolly ES, Jr., Rabinstein AA, Carhuapoma JR, et al. Guidelines for the management of aneurysmal subarachnoid hemorrhage: a guideline for healthcare professionals from the American Heart Association/american Stroke Association. Stroke. 2012;43(6):17111737. Chan M, Alaraj A, Calderon M, et al. Prediction of ventriculoperitoneal shunt dependency in patients with aneurysmal subarachnoid hemorrhage. J Neurosurg. 2009;110(1):44-49. Motiei-Langroudi R, Adeeb N, Foreman PM, et al. Predictors of Shunt Insertion in Aneurysmal Subarachnoid Hemorrhage. World Neurosurg. 2017;98:421-426. Wilson CD, Safavi-Abbasi S, Sun H, et al. Meta-analysis and systematic review of risk factors for shunt dependency after aneurysmal subarachnoid hemorrhage. J Neurosurg. 2017;126(2):586-595. Rincon F, Gordon E, Starke RM, et al. Predictors of long-term shunt-dependent hydrocephalus after aneurysmal subarachnoid hemorrhage. Clinical article. J Neurosurg. 2010;113(4):774-780. Czorlich P, Ricklefs F, Reitz M, et al. Impact of intraventricular hemorrhage measured by Graeb and LeRoux score on case fatality risk and chronic hydrocephalus in aneurysmal subarachnoid hemorrhage. Acta Neurochir (Wien). 2015;157(3):409-415. Oliveira DS, Tannus LR, Matheus AS, et al. [Evaluation of cardiovascular risk according to Framingham criteria in patients with type 2 diabetes]. Arq Bras Endocrinol Metabol. 2007;51(2):268-274. Dorai Z, Hynan LS, Kopitnik TA, Samson D. Factors related to hydrocephalus after aneurysmal subarachnoid hemorrhage. Neurosurgery. 2003;52(4):763-769; discussion 769-771. Diesing D, Wolf S, Sommerfeld J, Sarrafzadeh A, Vajkoczy P, Dengler NF. A novel score to predict shunt dependency after aneurysmal subarachnoid hemorrhage. J Neurosurg. 2017:1-7.
ACCEPTED MANUSCRIPT 17
18.
19.
20.
RI PT
17.
Jabbarli R, Bohrer AM, Pierscianek D, et al. The CHESS score: a simple tool for early prediction of shunt dependency after aneurysmal subarachnoid hemorrhage. Eur J Neurol. 2016;23(5):912-918. Wu Y, Green NL, Wrensch MR, Zhao S, Gupta N. Ventriculoperitoneal shunt complications in California: 1990 to 2000. Neurosurgery. 2007;61(3):557-562; discussion 562-553. Abla AA, Wilson DA, Williamson RW, et al. The relationship between ruptured aneurysm location, subarachnoid hemorrhage clot thickness, and incidence of radiographic or symptomatic vasospasm in patients enrolled in a prospective randomized controlled trial. J Neurosurg. 2014;120(2):391-397. Wilson DA, Nakaji P, Abla AA, et al. A simple and quantitative method to predict symptomatic vasospasm after subarachnoid hemorrhage based on computed tomography: beyond the Fisher scale. Neurosurgery. 2012;71(4):869-875. Wang YM, Lin YJ, Chuang MJ, et al. Predictors and outcomes of shunt-dependent hydrocephalus in patients with aneurysmal sub-arachnoid hemorrhage. BMC Surg. 2012;12:12.
SC
16.
M AN U
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321
AC C
EP
TE D
322
ACCEPTED MANUSCRIPT 18 323
Figure Legends
324 Figure 1. Study inclusion criteria.
326
EP
TE D
M AN U
SC
Figure 2. Receiver operating characteristic (ROC) curve.
AC C
327
RI PT
325
ACCEPTED MANUSCRIPT
Table 1. Patient characteristics Number (%)
RI PT
160 (66.1%) 82 (33.9%) 55.6 (±14.0) 115 (50.2%) 84 (42.9%) 40 (18.6%)
SC
59 (24.4%) 69 (28.5%) 50 (20.7%) 40 (16.5%) 24 (9.9%)
AC C
EP
TE D
M AN U
Variable Gender Female Male Age (± SD) HTN history1 Smoking history2 History of antiplatelet or anticoagulant use3 Hunt & Hess grade 1 2 3 4 5 GCS at admission4 3–8 9 – 12 13 – 15 Focal neurologic deficits at admission5 Modified Fisher CT scale grade at admission 0 1 2 3 4 Aneurysm location6 Anterior circulation Posterior circulation Aneurysm morphology7 Saccular Fusiform/Dissecting/Blister Aneurysm daughter sac8 Multiple aneurysms9 Hydrocephalus on admission10 EVD insertion during admission Treatment modality Clinical observation Microsurgical clipping Endovascular treatment
64 (28.2%) 12 (5.3%) 151 (66.5%) 30 (12.6%) 6 (2.5%) 9 (3.7%) 11 (4.6%) 67 (27.7%) 149 (61.6%) 131 (54.4%) 110 (45.6%) 218 (92.4%) 18 (7.6%) 57 (23.8%) 49 (20.9%) 101 (41.9%) 130 (53.7%) 10 (4.1%) 39 (16.1%) 192 (79.3%)
ACCEPTED MANUSCRIPT
1
Data missing for 13 patients. Data missing for 46 patients. 3 Data missing for 27 patients. 4 Data missing for 15 patients. 5 Data missing for 3 patients. 6 Data missing for 1 patient. 7 Data missing for 6 patients. 8 Data missing for 2 patients. 9 Data missing for 7 patients. 10 Data missing for 1 patient. 11 Data missing for 21 patients.
AC C
EP
TE D
2
1 (0.4%)
RI PT
182 (94.3%) 5 (2.6%) 5 (2.6%) 1 (0.5%)
SC
134 (60.6%) 19 (8.6%) 22 (10.0%) 26 (11.8%) 15 (6.8%) 5 (2.3%) 65 (26.9%) 40 (16.5%) 30 (12.4%) 54 (22.3%) 53 (21.9%)
M AN U
Open surgical and endovascular treatment Endovascular treatment modality Coil embolization Stent-assisted coil embolization Flow diversion Other modified Rankin Scale at discharge11 0 1 2 3 4 5 PS3 Score 0 1 2 3 4
ACCEPTED MANUSCRIPT
Table 2. Predictors of VP shunt insertion following aSAH VP shunt placement Yes: 54 (22.3%) No: 188 (77.7%)
EP
AC C
EVD insertion during hospital admission Treatment modality Clinical observation Microsurgical clipping Endovascular treatment Open surgical and endovascular
P-value
126 (67.0%) 62 (33.0%) 55.3 (±14.0) 91 (49.7%) 71 (45.5%)
0.47 0.76 0.14
11 (23.9%)
29 (17.2%)
0.29
SC
2 (3.7%) 13 (24.1%) 10 (18.5%) 18 (33.3%) 11 (20.4%)
0.57
RI PT
34 (63.0%) 20 (37.0%) 56.9 (±13.9) 24 (52.2%) 13 (32.5%)
57 (30.3%) 56 (29.8%) 40 (21.3%) 22 (11.7%) 13 (6.9%)
<0.001
27 (56.3%) 4(8.3%) 17 (35.4%) 7 (13.5%)
37 (20.7%) 8 (4.5%) 134 (74.9%) 23 (76.7%)
0 1 (1.9%) 0 5 (9.3%) 48 (88.9%) 31 (57.4%)
6 (3.2%) 8 (4.3%) 11 (5.9%) 62 (33.0%) 101 (53.7%) 70 (37.4%)
30 (55.6%) 24 (44.4%)
101 (54.0%) 86 (46.0%)
0.84
53 (98.2%)
77 (41.0%)
<0.001
1 (10.0%) 6 (15.4%) 47 (24.5%) 0
9 (90.0%) 33 (84.6%) 145 (75.5%) 1 (100.0%)
0.43
TE D
Gender Female Male Age HTN history Smoking history History of antiplatelet or anticoagulant use Hunt & Hess grade at admission 1 2 3 4 5 GCS at admission 3–8 9 – 12 13 - 15 Focal neurologic deficit at admission Modified Fisher CT scale grade at admission 0 1 2 3 4 Hydrocephalus on admission Aneurysm location Anterior circulation Posterior circulation
M AN U
Variable
<0.001 0.82
<0.001
0.01
ACCEPTED MANUSCRIPT
<0.001
0 0 5 (9.3%) 21 (38.9%) 28 (51.9%)
65 (34.6%) 40 (21.3%) 25 (13.3%) 33 (17.5%) 25 (13.3%)
<0.001
EP AC C
SC
RI PT
119 (67.6%) 12 (6.8%) 15 (8.5%) 18 (69.2%) 8 (53.3%) 4 (2.3%)
0 54 (100.0%)
TE D
2-4
15 (33.3%) 7 (15.6%) 7 (15.6%) 8 (17.8%) 7 (15.6%) 1 (2.2%)
M AN U
treatment modified Rankin Scale at discharge 0 1 2 3 4 5 PS3 score 0 1 2 3 4 Grouped PS3 system score 0–1
105 (55.9%) 83 (44.1%)
<0.001
ACCEPTED MANUSCRIPT
Table 3. Relationship between PS3 score and VP shunt insertion VP shunt placement No 188 (77.7%) 65 (34.6%) 40 (21.3%) 25 (13.3%) 33 (17.5%) 25 (13.3%)
AC C
EP
TE D
M AN U
SC
0 1 2 3 4
Yes 54 (22.3%) 0 (0%) 0 (0%) 5 (9.3%) 21 (38.9%) 28 (51.8%)
RI PT
PS3 score
ACCEPTED MANUSCRIPT
Table 4. Multivariable logistic regression analysis
0.92 (0.36-2.33) 51.7 (6.08-439.20)
EP AC C
0.70
0.87 <0.001
M AN U
1.92 (0.82-4.48) 1.74 (1.00-3.04)
RI PT
0.83 (0.32-2.13)
P-value 0.69 0.75 0.12 0.15
SC
OR (95% CI) 0.99 (0.96-1.02) 0.85 (0.32-2.25) 0.48 (0.18-1.23) 0.34 (0.08-1.47)
TE D
Variable Age ≥60 years Gender Smoking history Moderate or severe GCS at admission Aneurysm location in posterior circulation Acute hydrocephalus EVD placement during hospital admission Modified Fisher CT scale grade Hunt and Hess grade at admission
0.13 0.05
ACCEPTED MANUSCRIPT
Table 5. Predictive systems for VP shunt insertion
AC C
EP
TE D
M AN U
SC
RI PT
Scoring System and Criteria Points CHESS score Hunt and Hess grade ≥ 1 1 Aneurysm location in posterior circulation 1 Acute hydrocephalus 4 Intraventricular hemorrhage 1 Cerebral infarction on CT imaging 1 Total possible points 8 * BNI Score No visible SAH 1 ≤5 mm 2 6-10 mm 3 11-15 mm 4 16-20 mm 5 Total possible points 5 SDASH score Hunt and Hess grade ≥ 4 1 * BNI score ≥ 3 1 Acute hydrocephalus 2 Total possible points 4 PS3 score Hunt and Hess grade ≥ 4 1 Modified Fisher CT score = 4 1 EVD insertion 2 Total possible points 4 * The BNI score assesses the thickness of the subarachnoid blood clot.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Highlights: Hydrocephalus is a frequently encountered complication in the context of aSAH.
•
The recently proposed PS3 system estimates the risk of VP shunt insertion post-aSAH.
•
We externally validate this system in a cohort of 242 patients and find an area under the
RI PT
•
ROC of 0.845. •
The PS3 score reliability predicted shunt-dependent chronic hydrocephalus (OR: 3.36;
The PS3 system more accurately predicted shunt-dependent chronic hydrocephalus in
M AN U
patients with aSAH, as compared to all other such systems in the neurosurgical literature
EP
TE D
such as the CHESS, BNI, and SDASH systems.
AC C
•
SC
95% CI: 2.31-4.89; p <0.001).
ACCEPTED MANUSCRIPT
Abbreviations: Aneurysmal subarachnoid hemorrhage (aSAH)
•
Ventriculoperitoneal (VP)
•
Computed tomography (CT)
•
Post-subarachnoid shunt scoring system (PS3)
•
Receiver operating characteristic (ROC)
•
Glasgow Coma Scale (GCS)
•
External ventricular drain (EVD)
•
Modified Ranking Scale (mRS)
•
Standard deviation (SD)
•
Confidence interval (CI)
•
Failure risk index (FRI)
•
Chronic Hydrocephalus Ensuing from SAH Score (CHESS)
•
Shunt Dependency in aSAH (SDASH)
•
Barrow Neurological Institute (BNI)
•
Intraventricular hemorrhage (ICH)
•
Cerebrospinal fluid (CSF)
SC M AN U
TE D
EP
AC C
RI PT
•
S1878-8750(17)31656-X
DOI:
10.1016/j.wneu.2017.09.140
Reference:
WNEU 6584
To appear in:
World Neurosurgery
Received Date: 22 August 2017 Revised Date:
19 September 2017
Accepted Date: 20 September 2017
Please cite this article as: Gupta R, Ascanio LC, Enriquez-Marulanda A, Griessenauer CJ, Chinnadurai A, Jhun R, Alturki A, Ogilvy CS, Thomas AJ, Moore JM, Validation of a Predictive Scoring System for Ventriculoperitoneal Shunt Insertion following Aneurysmal Subarachnoid Hemorrhage, World Neurosurgery (2017), doi: 10.1016/j.wneu.2017.09.140. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1 Validation of a Predictive Scoring System for Ventriculoperitoneal Shunt Insertion
2
following Aneurysmal Subarachnoid Hemorrhage
3
Raghav Gupta, B.S., * Luis C. Ascanio, M.D., * Alejandro Enriquez-Marulanda, M.D., Christoph
4
J. Griessenauer, M.D., Anu Chinnadurai, Ray Jhun, B.S., Abdulrahman Alturki, M.B.B.S.,
5
M.Sc., FRCSC, Christopher S. Ogilvy, M.D., Ajith J. Thomas, M.D., Justin M. Moore, M.D.,
6
Ph.D.
7
Neurosurgical Service, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston,
8
Massachusetts, USA
9
*These authors contributed equally to this study.
M AN U
SC
RI PT
1
10
Abbreviations: Attached on a separate sheet.
12
Keywords: chronic hydrocephalus; venticuloperitonal shunt; subarachnoid hemorrhage; scoring
13
system
14
TE D
11
Corresponding Author
16
Justin M. Moore, M.D., Ph.D.
17
Neurosurgical Service
18
Beth Israel Deaconess Medical Center
19
Harvard Medical School
20
110 Francis Street
21
Boston, MA, 02215
22
Email: [email protected]
23
Phone: 617-632-7246
24
AC C
EP
15
ACCEPTED MANUSCRIPT 2 25
Abstract
26 Introduction
28
Hydrocephalus is a frequently encountered complication in the context of aneurysmal
29
subarachnoid hemorrhage (aSAH). Here, we performed an external validation of the recently
30
proposed Post-Subarachnoid Shunt Scoring (PS3) system, which aims to stratify patients
31
presenting with aSAH based on their relative risk of requiring ventriculoperitoneal (VP) shunt
32
insertion.
SC
RI PT
27
M AN U
33 Materials and Methods
35
A retrospective review of all patients presenting with aSAH to our institution between July 2007
36
and December 2016, who underwent CT imaging at the time of hospital admission, was
37
performed.
38
TE D
34
Results
40
A total of 242 patients (66.1% female) with aSAH were included in the analysis with a mean age
41
of 55.6 years. Sixty-four (26.4%) patients had a Hunt and Hess grade of 4 or 5 on admission. An
42
external ventricular drain (EVD) was placed in 130 (53.7%) patients during the hospital
43
admission. EVD placement was found to correlate with an increased rate of VP shunt placement
44
(p<0.001), and a trend towards an association between a high Hunt and Hess grade and VP shunt
45
placement, was observed (p=0.05). The area under the ROC for the PS3 scoring system was
46
found to be 0.845. The system reliably predicted shunt-dependent chronic hydrocephalus in our
47
patient cohort (Odds ratio [OR]: 3.36; 95% confidence interval [CI]: 2.31-4.89; p <0.001).
AC C
EP
39
ACCEPTED MANUSCRIPT 3 48 Conclusion
50
Data from this study validated the previously proposed PS3 system, which was found to more
51
accurately predict shunt-dependent chronic hydrocephalus in patients with aSAH, as compared to
52
other such systems in the neurosurgical literature such as the CHESS, BNI, and SDASH systems.
RI PT
49
53 Word Count: 249
SC
54 55
AC C
EP
TE D
M AN U
56
ACCEPTED MANUSCRIPT 4 57
Introduction
58 Nearly 80% of nontraumatic subarachnoid hemorrhage (SAH) cases are attributable to ruptured
60
intracranial aneurysms.1 The resulting case fatality rate is greater than 50%, and patients who
61
survive often suffer from cognitive deficits and a subsequent deterioration in quality-of-life.2,3
62
Hydrocephalus is a frequently encountered complication in the context of SAH, and can occur in
63
up to 30% of patients.4 It is classified as being either acute (0-3 days post-SAH), subacute (4-13
64
days post-SAH), or chronic (≥14 days post-SAH).5 The incidence of chronic hydrocephalus after
65
SAH, which is typically managed through diversion of cerebrospinal fluid (CSF) flow via
66
ventriculoperitoneal (VP) shunt insertion, is reported to be between 6.0 and 48.0%.6,7
M AN U
SC
RI PT
59
67
Prior studies evaluating factors associated with VP shunt placement have found that a larger third
69
ventricular diameter at admission, history of sympathomimetic drug use, hyponatremia, high
70
Fisher computed tomography (CT) grade, low initial Glasgow Coma Scale (GCS) score, a higher
71
Hunt and Hess grade at admission, female gender, aneurysm location within the posterior
72
circulation, aneurysm size, and age, each predict the need for subsequent shunt insertion.6,8
73
However, in a recent study published by our group., a simpler scoring system based on external
74
ventricular drain (EVD) insertion, a high (≥4) Hunt and Hess grade at admission, and a modified
75
Fisher CT grade of 4 was proposed based on patients enrolled in the Cerebral Aneurysm Renin
76
Angiotensin System (CARAS) study.9 It was found to accurately predict the need for shunt
77
placement in a cohort of 149 patients presenting with aneurysmal SAH (aSAH). In the present
78
study, we aimed to validate this PS3 (Post-Subarachnoid Shunt Scoring) system using data from
AC C
EP
TE D
68
ACCEPTED MANUSCRIPT 5 79
a separate cohort of patients who presented with aSAH to a major academic center in the U.S.,
80
over an 8-year period, who had not been a part of the original CARAS study.
AC C
EP
TE D
M AN U
SC
RI PT
81
ACCEPTED MANUSCRIPT 6 82
Materials and Methods
83 Patient Selection
85
Institutional Review Board approval was obtained prior to the commencement this study. A
86
retrospective review of all consecutive patients presenting with SAH, between July 2007 and
87
December 2016, to an academic neurosurgical department in the U.S., was performed. Patients
88
were included if the etiology of their SAH was aneurysmal, and they had received a CT scan
89
upon being admitted. Patients were excluded if they presented with either non-traumatic non-
90
aneurysmal SAH, expired prior to initial hospital discharge, or if they had not undergone CT
91
imaging at the time of hospital admission (Figure 1).
M AN U
SC
RI PT
84
92
The variables collected for each patient included: patient demographics (age, gender, smoking
94
status, history of hypertension, history of antiplatelet/anticoagulant use), initial clinical
95
presentation (Hunt and Hess grade on admission prior to EVD insertion, Glasgow Coma Scale
96
(GCS) score on admission, presence of focal neurological deficits on admission, and Modified
97
Fisher CT scale grade on admission), aneurysm characteristics (aneurysm location, morphology,
98
presence of a daughter sac), EVD placement within the same hospital admission, treatment
99
modality, and modified Rankin Scale (mRS) score at discharge. The Hunt and Hess scale grade,
100
modified Fisher CT scale grade, and history of EVD insertion, were used to tabulate a score for
101
each patient as per the PS3 system.9
103
EP
AC C
102
TE D
93
External Ventricular Drain and Ventriculoperitoneal Shunt Placement Protocol
ACCEPTED MANUSCRIPT 7 The protocol at our institution is to place an EVD in all patients presenting with acute
105
hydrocephalus (confirmed via radiographic and/or clinical examination) or in patients with
106
suspected increased intracranial pressure (ICP) based on physical examination. The decision to
107
insert a VP shunt rather than to perform additional EVD clamp trials was based on multiple
108
criteria including the number of clamp trials that had been already performed, the severity of the
109
hydrocephalus as determined via CT imaging, and the rise in ICP once the EVD had been
110
clamped. Other factors including a high Hunt and Hess grade and modified Fisher CT grade at
111
admission, were used to guide decision-making. The clinical gestalt of the two cerebrovascular
112
neurosurgeons involved in the management of these patients, however, often heavily influenced
113
the decision to insert a shunt or to perform additional clamp trials.
M AN U
SC
RI PT
104
114 Statistical Analyses
116
The Chi-square test was used for the analysis of categorical variables. The t-test was used for
117
analyzing normally distributed continuous variables. A mean and standard deviation (SD) were
118
reported for normally distributed variables. The scoring system was validated via development of
119
a receiver operating characteristic (ROC) curve. The area under the curve was then calculated.
120
The Hosmer-Lemeshow test was performed to assess goodness of fit of the univariable logistic
121
regression model that was developed; goodness of fit was defined by a p-value >0.05. Statistics
122
were performed using the STATA software (StataCorp, Texas, U.S.). Significance was defined
123
with a p-value <0.05.
125
EP
AC C
124
TE D
115
ACCEPTED MANUSCRIPT 8 126
Results
127 Patients Characteristics
129
A total of 242 patients presented with aSAH to our institution and underwent CT imaging upon
130
hospital admission, within this time period. The mean age of these patients was 55.6 years
131
(Standard deviation[SD] ± 14.0 years). A history of hypertension or smoking was noted in 115
132
(50.2%) and 84 (42.9%) patients, respectively (Table 1). A majority (54.4%) of patients had
133
aneurysms located within the anterior circulation and of saccular morphology (92.4%)
SC
RI PT
128
M AN U
134 Clinical Presentation
136
Sixty-four (26.4%) patients had a Hunt and Hess grade of 4 or 5 on admission. A majority (149;
137
61.6%) of patients had a modified Fisher CT grade 4 hemorrhage. Thirty (12.6%) patients
138
presented with focal neurological deficits. Acute hydrocephalus was detected in 101 (41.9%)
139
patients. An EVD was placed in 130 (53.7%) patients. Fifty-four (22.3%) patients underwent VP
140
shunt placement for chronic hydrocephalus. Nine of these shunts were inserted in a separate
141
hospital admission due to the onset of delayed hydrocephalus while the remaining shunts were
142
inserted in the same admission period as the initial hemorrhage.
EP
AC C
143
TE D
135
144
Predictors of Ventriculoperitoneal Shunt Insertion
145
No differences in gender (p=0.57), age (p=0.47), history of hypertension (p=0.76), smoking
146
status (p=0.14), previous antiplatelet/anticoagulant use (p=0.29), and presence of neurologic
147
deficits at the time of admission (p=0.82), were noted in patients who did and did not undergo
148
VP shunt placement (Table 2). Aneurysm treatment modality was not found to be a predictor of
ACCEPTED MANUSCRIPT 9 VP shunt placement (p=0.43). Conversely, a higher Hunt and Hess grade at admission (p<0.001),
150
a lower GCS at admission (p<0.001), a higher modified Fisher CT grade hemorrhage (p<0.001),
151
EVD insertion during hospital admission (p<0.001), a higher mRS at discharge (p<0.001), and a
152
higher score on the PS3 system (p<0.001) were found to predict the need for VP shunt insertion.
153
Twenty-eight (51.8%) patients with a PS3 score of 4 underwent VP shunt placement as
154
compared to zero patients who had a PS3 score of 1 (Table 3). Patients with a PS3 score of 3 or 4
155
were significantly more likely to have a VP shunt inserted than patients with a score of 0, 1, or 2
156
(OR: 17.66; p<0.001).
157
M AN U
SC
RI PT
149
Multivariable Logistic Regression, Univariable Logistic Regression, and Receiver Operating
159
Characteristic Curve Analyses
160
On multivariable regression analysis, EVD placement (p<0.001) was found to correlate with an
161
increased rate of VP shunt placement (Table 4). A trend towards an association between high
162
Hunt and Hess grade at admission and VP shunt placement was observed (p=0.05). The area
163
under the ROC for the PS3 scoring system was found to be 0.845 (Figure 2). On univariable
164
logistic regression analysis, the PS3 score reliability predicted shunt-dependent chronic
165
hydrocephalus in our patient cohort (Odds ratio [OR]: 3.36; 95% confidence interval [CI]: 2.31-
166
4.89; p <0.001). The Hosmer-Lemeshow test demonstrated goodness of fit of the PS3 scoring
167
system (p=0.11).
EP
AC C
168
TE D
158
ACCEPTED MANUSCRIPT 10 169
Discussion
170 Background
172
Chronic hydrocephalus is a clinically important complication of aSAH, with up to 48% of
173
patients in whom EVDs are placed requiring VP shunt insertion for management of this
174
condition.6,7 Several studies in the neurosurgical literature identifying factors predictive of the
175
need for shunt insertion following aSAH have been published, albeit with significant
176
heterogeneity in regards to the individual factors that they have identified.5,6,8-14 A recent
177
systematic review and meta-analysis of these studies by Wilson et al., identified higher Fisher
178
grade (3/4 vs 1/2), presentation with acute hydrocephalus, in-hospital complications (including
179
pneumonia, meningitis, vasospasm, and ischemic strokes), intraventricular hemorrhage, high
180
Hunt and Hess grade, older age, female gender, and aneurysm location within the posterior
181
circulation as risk factors for shunt insertion.10 Of these factors, presentation with acute
182
hydrocephalus (OR: 5.67; 95% CI: 3.96-8.12) and a higher Fisher grade on CT imaging (OR:
183
7.74; CI: 4.47-13.41) were the two strongest predictors of the need for VP shunt placement.9
SC
M AN U
TE D
EP
184
RI PT
171
Previous Predictive Algorithms for Ventriculoperitoneal Shunt Insertion
186
Development of a simple system which could stratify patients presenting with aSAH based on
187
the risk of requiring surgical shunt placement, is essential. For high-risk patients, such a system
188
could potentially guide neurosurgeons towards replacing ventricular catheters with permanent
189
shunts more rapidly, leading to a decreased incidence of nosocomial EVD infections, fewer
190
clamp trial failures, shortened hospitals stays, and an overall decrease in treatment costs.15,16
191
Conversely, for patients at low risk of requiring VP shunt placement, a more aggressive EVD
AC C
185
ACCEPTED MANUSCRIPT 11 192
weaning process could be utilized as shunt placement is associated with several complications
193
including infections, shunt mechanical dysfunction, and/or obstruction.17
194 However, few such scoring systems exist within the neurosurgical literature. These include the
196
failure risk index (FRI) score,8 the Chronic Hydrocephalus Ensuing from SAH Score (CHESS),16
197
and the Shunt Dependency in aSAH (SDASH) score (Table 5).15 The FRI, however, has not been
198
widely used in the clinical setting.15 The Barrow Neurological Institute (BNI) scale,18,19 which
199
was originally developed to predict the incidence of symptomatic vasospasm following aSAH as
200
a function of the thickness of the blood clot, was also predictive of shunt-dependent chronic
201
hydrocephalus (area under ROC: 0.649).15 In the present study, we found the PS3 score more
202
accurately (OR: 3.36; 95% CI: 2.31-4.89; p <0.001; area under ROC 0.845) predicted the need
203
for shunt placement in a large cohort of patients, than either the SDASH (OR: 2.50; 95% CI:
204
1.86-3.35; p<0.001; area under ROC: 0.785), the CHESS (OR: 1.53; 95% CI 1.33-1.78; p<0.001;
205
area under ROC: 0.769), or the BNI (OR: 2.02; 95% CI: 1.45-2.82; p<0.001; area under ROC:
206
0.649) scoring systems.15 EVD placement in the same hospital admission was found to
207
independently predict the need for shunt placement even after age, gender, GCS score,
208
presentation with acute hydrocephalus, and aneurysm location were controlled for in a
209
multivariable regression model. A trend towards an association between a higher Hunt and Hess
210
grade at admission and shunt placement was observed as well (p=0.05). Both parameters are
211
components of the PS3 scoring system.
213
SC
M AN U
TE D
EP
AC C
212
RI PT
195
Clinical Utility of the PS3 Scoring System
ACCEPTED MANUSCRIPT 12 Adoption of a scoring system within the clinical setting for the rapid stratification of patients is
215
predicated on the simplicity of the system. The PS3 scoring system contains three factors, two of
216
which (Hunt and Hess & Modified Fisher CT score) are routinely assessed in patients presenting
217
with aSAH, and have been each independently found to predict the need for shunt insertion.10,20
218
The third (EVD insertion) is a binary variable (yes or no) which can be easily and rapidly
219
recorded. The modified Fisher CT scale assesses whether there is an intraventricular hemorrhage
220
(IVH) component, and is a variable listed separately in the more complex CHESS scoring
221
system. As obstruction of cerebrospinal fluid (CSF) flow has been proposed as a possible basis
222
for the development of hydrocephalus,14 the presence of IVH could serve as the physiological
223
mechanism by which this occurs. Additionally, the modified Fisher scale score is a more detailed
224
variant of the Fisher scale score, which was found to be the strongest predictor of shunt-
225
dependent hydrocephalus in Wilson et al.’s meta-analysis.10
M AN U
SC
RI PT
214
TE D
226 Limitations
228
This study is limited by its retrospective design. However, the PS3 system was validated in the
229
current study using a large cohort of patients who presented with aSAH to a different institution
230
than the one where the system was originally proposed. The decision to place a VP shunt was
231
made on a per-patient basis by the attending physician, and different neurosurgeons may have
232
various thresholds for shunt insertion which are difficult to assess.
AC C
233
EP
227
ACCEPTED MANUSCRIPT 13 234
Conclusion
235 Chronic hydrocephalus is a clinically important condition with a significant proportion of
237
patients who present with aSAH requiring eventual surgical shunt insertion. Few scoring systems
238
exist in the neurosurgical literature which can rapidly stratify patients based on their risk of
239
requiring VP shunt placement. The PS3 scoring system was recently proposed to address this,
240
and is based on Hunt and Hess grade, modified Fisher CT score, and EVD insertion. In the
241
present study, we performed an external validation of the PS3 score, and found it to more
242
accurately predict shunt-dependent chronic hydrocephalus than previous such scoring systems
243
such as the CHESS, BNI, and SDASH systems.
M AN U
SC
RI PT
236
AC C
EP
TE D
244
ACCEPTED MANUSCRIPT 14 245
Acknowledgments
246 Conflict of Interest Statement
248
The authors have no personal, financial, or institutional interest with regards to the authorship
249
and/or publication of this manuscript.
RI PT
247
250 Funding Statement
252
This research did not receive any specific grant from funding agencies in the public, commercial,
253
or not-for-profit sectors.
M AN U
SC
251
254
AC C
EP
TE D
255
ACCEPTED MANUSCRIPT 15
257 258 259 260
Additional Author Affiliations 1. C.J.G. Department of Neurosurgery, Geisinger Medical Center, Danville, Pennsylvania, USA 2. A.A. Department of Neurosurgery, National Neuroscience Institute, King Fahad Medical
RI PT
256
City, Riyadh, Saudi Arabia
AC C
EP
TE D
M AN U
SC
261
ACCEPTED MANUSCRIPT 16 262
References
263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304
1.
7.
8.
9. 10.
11.
12.
13.
14.
15.
RI PT
SC
6.
M AN U
5.
TE D
4.
EP
3.
AC C
2.
Suarez JI, Tarr RW, Selman WR. Aneurysmal subarachnoid hemorrhage. N Engl J Med. 2006;354(4):387-396. Mayer SA, Kreiter KT, Copeland D, et al. Global and domain-specific cognitive impairment and outcome after subarachnoid hemorrhage. Neurology. 2002;59(11):17501758. Hackett ML, Anderson CS. Health outcomes 1 year after subarachnoid hemorrhage: An international population-based study. The Australian Cooperative Research on Subarachnoid Hemorrhage Study Group. Neurology. 2000;55(5):658-662. Germanwala AV, Huang J, Tamargo RJ. Hydrocephalus after aneurysmal subarachnoid hemorrhage. Neurosurg Clin N Am. 2010;21(2):263-270. Vale FL, Bradley EL, Fisher WS, 3rd. The relationship of subarachnoid hemorrhage and the need for postoperative shunting. J Neurosurg. 1997;86(3):462-466. O'Kelly CJ, Kulkarni AV, Austin PC, Urbach D, Wallace MC. Shunt-dependent hydrocephalus after aneurysmal subarachnoid hemorrhage: incidence, predictors, and revision rates. Clinical article. J Neurosurg. 2009;111(5):1029-1035. Connolly ES, Jr., Rabinstein AA, Carhuapoma JR, et al. Guidelines for the management of aneurysmal subarachnoid hemorrhage: a guideline for healthcare professionals from the American Heart Association/american Stroke Association. Stroke. 2012;43(6):17111737. Chan M, Alaraj A, Calderon M, et al. Prediction of ventriculoperitoneal shunt dependency in patients with aneurysmal subarachnoid hemorrhage. J Neurosurg. 2009;110(1):44-49. Motiei-Langroudi R, Adeeb N, Foreman PM, et al. Predictors of Shunt Insertion in Aneurysmal Subarachnoid Hemorrhage. World Neurosurg. 2017;98:421-426. Wilson CD, Safavi-Abbasi S, Sun H, et al. Meta-analysis and systematic review of risk factors for shunt dependency after aneurysmal subarachnoid hemorrhage. J Neurosurg. 2017;126(2):586-595. Rincon F, Gordon E, Starke RM, et al. Predictors of long-term shunt-dependent hydrocephalus after aneurysmal subarachnoid hemorrhage. Clinical article. J Neurosurg. 2010;113(4):774-780. Czorlich P, Ricklefs F, Reitz M, et al. Impact of intraventricular hemorrhage measured by Graeb and LeRoux score on case fatality risk and chronic hydrocephalus in aneurysmal subarachnoid hemorrhage. Acta Neurochir (Wien). 2015;157(3):409-415. Oliveira DS, Tannus LR, Matheus AS, et al. [Evaluation of cardiovascular risk according to Framingham criteria in patients with type 2 diabetes]. Arq Bras Endocrinol Metabol. 2007;51(2):268-274. Dorai Z, Hynan LS, Kopitnik TA, Samson D. Factors related to hydrocephalus after aneurysmal subarachnoid hemorrhage. Neurosurgery. 2003;52(4):763-769; discussion 769-771. Diesing D, Wolf S, Sommerfeld J, Sarrafzadeh A, Vajkoczy P, Dengler NF. A novel score to predict shunt dependency after aneurysmal subarachnoid hemorrhage. J Neurosurg. 2017:1-7.
ACCEPTED MANUSCRIPT 17
18.
19.
20.
RI PT
17.
Jabbarli R, Bohrer AM, Pierscianek D, et al. The CHESS score: a simple tool for early prediction of shunt dependency after aneurysmal subarachnoid hemorrhage. Eur J Neurol. 2016;23(5):912-918. Wu Y, Green NL, Wrensch MR, Zhao S, Gupta N. Ventriculoperitoneal shunt complications in California: 1990 to 2000. Neurosurgery. 2007;61(3):557-562; discussion 562-553. Abla AA, Wilson DA, Williamson RW, et al. The relationship between ruptured aneurysm location, subarachnoid hemorrhage clot thickness, and incidence of radiographic or symptomatic vasospasm in patients enrolled in a prospective randomized controlled trial. J Neurosurg. 2014;120(2):391-397. Wilson DA, Nakaji P, Abla AA, et al. A simple and quantitative method to predict symptomatic vasospasm after subarachnoid hemorrhage based on computed tomography: beyond the Fisher scale. Neurosurgery. 2012;71(4):869-875. Wang YM, Lin YJ, Chuang MJ, et al. Predictors and outcomes of shunt-dependent hydrocephalus in patients with aneurysmal sub-arachnoid hemorrhage. BMC Surg. 2012;12:12.
SC
16.
M AN U
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321
AC C
EP
TE D
322
ACCEPTED MANUSCRIPT 18 323
Figure Legends
324 Figure 1. Study inclusion criteria.
326
EP
TE D
M AN U
SC
Figure 2. Receiver operating characteristic (ROC) curve.
AC C
327
RI PT
325
ACCEPTED MANUSCRIPT
Table 1. Patient characteristics Number (%)
RI PT
160 (66.1%) 82 (33.9%) 55.6 (±14.0) 115 (50.2%) 84 (42.9%) 40 (18.6%)
SC
59 (24.4%) 69 (28.5%) 50 (20.7%) 40 (16.5%) 24 (9.9%)
AC C
EP
TE D
M AN U
Variable Gender Female Male Age (± SD) HTN history1 Smoking history2 History of antiplatelet or anticoagulant use3 Hunt & Hess grade 1 2 3 4 5 GCS at admission4 3–8 9 – 12 13 – 15 Focal neurologic deficits at admission5 Modified Fisher CT scale grade at admission 0 1 2 3 4 Aneurysm location6 Anterior circulation Posterior circulation Aneurysm morphology7 Saccular Fusiform/Dissecting/Blister Aneurysm daughter sac8 Multiple aneurysms9 Hydrocephalus on admission10 EVD insertion during admission Treatment modality Clinical observation Microsurgical clipping Endovascular treatment
64 (28.2%) 12 (5.3%) 151 (66.5%) 30 (12.6%) 6 (2.5%) 9 (3.7%) 11 (4.6%) 67 (27.7%) 149 (61.6%) 131 (54.4%) 110 (45.6%) 218 (92.4%) 18 (7.6%) 57 (23.8%) 49 (20.9%) 101 (41.9%) 130 (53.7%) 10 (4.1%) 39 (16.1%) 192 (79.3%)
ACCEPTED MANUSCRIPT
1
Data missing for 13 patients. Data missing for 46 patients. 3 Data missing for 27 patients. 4 Data missing for 15 patients. 5 Data missing for 3 patients. 6 Data missing for 1 patient. 7 Data missing for 6 patients. 8 Data missing for 2 patients. 9 Data missing for 7 patients. 10 Data missing for 1 patient. 11 Data missing for 21 patients.
AC C
EP
TE D
2
1 (0.4%)
RI PT
182 (94.3%) 5 (2.6%) 5 (2.6%) 1 (0.5%)
SC
134 (60.6%) 19 (8.6%) 22 (10.0%) 26 (11.8%) 15 (6.8%) 5 (2.3%) 65 (26.9%) 40 (16.5%) 30 (12.4%) 54 (22.3%) 53 (21.9%)
M AN U
Open surgical and endovascular treatment Endovascular treatment modality Coil embolization Stent-assisted coil embolization Flow diversion Other modified Rankin Scale at discharge11 0 1 2 3 4 5 PS3 Score 0 1 2 3 4
ACCEPTED MANUSCRIPT
Table 2. Predictors of VP shunt insertion following aSAH VP shunt placement Yes: 54 (22.3%) No: 188 (77.7%)
EP
AC C
EVD insertion during hospital admission Treatment modality Clinical observation Microsurgical clipping Endovascular treatment Open surgical and endovascular
P-value
126 (67.0%) 62 (33.0%) 55.3 (±14.0) 91 (49.7%) 71 (45.5%)
0.47 0.76 0.14
11 (23.9%)
29 (17.2%)
0.29
SC
2 (3.7%) 13 (24.1%) 10 (18.5%) 18 (33.3%) 11 (20.4%)
0.57
RI PT
34 (63.0%) 20 (37.0%) 56.9 (±13.9) 24 (52.2%) 13 (32.5%)
57 (30.3%) 56 (29.8%) 40 (21.3%) 22 (11.7%) 13 (6.9%)
<0.001
27 (56.3%) 4(8.3%) 17 (35.4%) 7 (13.5%)
37 (20.7%) 8 (4.5%) 134 (74.9%) 23 (76.7%)
0 1 (1.9%) 0 5 (9.3%) 48 (88.9%) 31 (57.4%)
6 (3.2%) 8 (4.3%) 11 (5.9%) 62 (33.0%) 101 (53.7%) 70 (37.4%)
30 (55.6%) 24 (44.4%)
101 (54.0%) 86 (46.0%)
0.84
53 (98.2%)
77 (41.0%)
<0.001
1 (10.0%) 6 (15.4%) 47 (24.5%) 0
9 (90.0%) 33 (84.6%) 145 (75.5%) 1 (100.0%)
0.43
TE D
Gender Female Male Age HTN history Smoking history History of antiplatelet or anticoagulant use Hunt & Hess grade at admission 1 2 3 4 5 GCS at admission 3–8 9 – 12 13 - 15 Focal neurologic deficit at admission Modified Fisher CT scale grade at admission 0 1 2 3 4 Hydrocephalus on admission Aneurysm location Anterior circulation Posterior circulation
M AN U
Variable
<0.001 0.82
<0.001
0.01
ACCEPTED MANUSCRIPT
<0.001
0 0 5 (9.3%) 21 (38.9%) 28 (51.9%)
65 (34.6%) 40 (21.3%) 25 (13.3%) 33 (17.5%) 25 (13.3%)
<0.001
EP AC C
SC
RI PT
119 (67.6%) 12 (6.8%) 15 (8.5%) 18 (69.2%) 8 (53.3%) 4 (2.3%)
0 54 (100.0%)
TE D
2-4
15 (33.3%) 7 (15.6%) 7 (15.6%) 8 (17.8%) 7 (15.6%) 1 (2.2%)
M AN U
treatment modified Rankin Scale at discharge 0 1 2 3 4 5 PS3 score 0 1 2 3 4 Grouped PS3 system score 0–1
105 (55.9%) 83 (44.1%)
<0.001
ACCEPTED MANUSCRIPT
Table 3. Relationship between PS3 score and VP shunt insertion VP shunt placement No 188 (77.7%) 65 (34.6%) 40 (21.3%) 25 (13.3%) 33 (17.5%) 25 (13.3%)
AC C
EP
TE D
M AN U
SC
0 1 2 3 4
Yes 54 (22.3%) 0 (0%) 0 (0%) 5 (9.3%) 21 (38.9%) 28 (51.8%)
RI PT
PS3 score
ACCEPTED MANUSCRIPT
Table 4. Multivariable logistic regression analysis
0.92 (0.36-2.33) 51.7 (6.08-439.20)
EP AC C
0.70
0.87 <0.001
M AN U
1.92 (0.82-4.48) 1.74 (1.00-3.04)
RI PT
0.83 (0.32-2.13)
P-value 0.69 0.75 0.12 0.15
SC
OR (95% CI) 0.99 (0.96-1.02) 0.85 (0.32-2.25) 0.48 (0.18-1.23) 0.34 (0.08-1.47)
TE D
Variable Age ≥60 years Gender Smoking history Moderate or severe GCS at admission Aneurysm location in posterior circulation Acute hydrocephalus EVD placement during hospital admission Modified Fisher CT scale grade Hunt and Hess grade at admission
0.13 0.05
ACCEPTED MANUSCRIPT
Table 5. Predictive systems for VP shunt insertion
AC C
EP
TE D
M AN U
SC
RI PT
Scoring System and Criteria Points CHESS score Hunt and Hess grade ≥ 1 1 Aneurysm location in posterior circulation 1 Acute hydrocephalus 4 Intraventricular hemorrhage 1 Cerebral infarction on CT imaging 1 Total possible points 8 * BNI Score No visible SAH 1 ≤5 mm 2 6-10 mm 3 11-15 mm 4 16-20 mm 5 Total possible points 5 SDASH score Hunt and Hess grade ≥ 4 1 * BNI score ≥ 3 1 Acute hydrocephalus 2 Total possible points 4 PS3 score Hunt and Hess grade ≥ 4 1 Modified Fisher CT score = 4 1 EVD insertion 2 Total possible points 4 * The BNI score assesses the thickness of the subarachnoid blood clot.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Highlights: Hydrocephalus is a frequently encountered complication in the context of aSAH.
•
The recently proposed PS3 system estimates the risk of VP shunt insertion post-aSAH.
•
We externally validate this system in a cohort of 242 patients and find an area under the
RI PT
•
ROC of 0.845. •
The PS3 score reliability predicted shunt-dependent chronic hydrocephalus (OR: 3.36;
The PS3 system more accurately predicted shunt-dependent chronic hydrocephalus in
M AN U
patients with aSAH, as compared to all other such systems in the neurosurgical literature
EP
TE D
such as the CHESS, BNI, and SDASH systems.
AC C
•
SC
95% CI: 2.31-4.89; p <0.001).
ACCEPTED MANUSCRIPT
Abbreviations: Aneurysmal subarachnoid hemorrhage (aSAH)
•
Ventriculoperitoneal (VP)
•
Computed tomography (CT)
•
Post-subarachnoid shunt scoring system (PS3)
•
Receiver operating characteristic (ROC)
•
Glasgow Coma Scale (GCS)
•
External ventricular drain (EVD)
•
Modified Ranking Scale (mRS)
•
Standard deviation (SD)
•
Confidence interval (CI)
•
Failure risk index (FRI)
•
Chronic Hydrocephalus Ensuing from SAH Score (CHESS)
•
Shunt Dependency in aSAH (SDASH)
•
Barrow Neurological Institute (BNI)
•
Intraventricular hemorrhage (ICH)
•
Cerebrospinal fluid (CSF)
SC M AN U
TE D
EP
AC C
RI PT
•