Neurocrit Care DOI 10.1007/s12028-015-0145-6
ORIGINAL ARTICLE
Acutely Trapped Ventricle: Clinical Significance and Benefit from Surgical Decompression Gabriel L. Pagani-Este´vez1 • Philippe Couillard1 • Giuseppe Lanzino2 Eelco F. M. Wijdicks1 • Alejandro A. Rabinstein1
•
Ó Springer Science+Business Media New York 2015
Abstract Background Focal ventricular obstruction—trapped ventricle—results in cerebrospinal fluid accumulation, mass effect and possible clinical deterioration. There are no systematic studies on the benefit of surgical decompression in adults. Methods We reviewed patients admitted with acutely trapped ventricle on brain imaging to assess their prognosis and the effect of surgical intervention on 30-day mortality. Results Of the 392 patients with trapped ventricle, the most common causes were brain tumor (45 %), intracerebral hemorrhage (ICH) (20 %), and subdural hematoma (SDH) (14 %). Lateral ventricle trapping accounted for 97 % of cases. Two hundred and twenty-one patients (56 %) received a surgical intervention for trapped ventricle or its causes; 126 (83 %) were treated with craniotomy, 26 (17 %) with craniectomy, 30 (14 %) with external ventricular drain (EVD) alone, 23 (10 %) with ventriculoperitoneal shunt alone, and 16 (7 %) with endoscopic fenestration of the septum pellucidum. Surgical intervention was associated with mortality reduction from 95 % (n = 54) to 48 % (n = 11) in the ICH group, from 47 % (n = 27) to 12 % (n = 15) in the tumor group and from 90 % (n = 18) to 20 % (n = 7) in the SDH group (p < 0.001 for all comparisons). Univariate logistic analysis showed that surgical intervention and tumor etiology were associated with decreased mortality while age, ICH & Alejandro A. Rabinstein [email protected] 1
Department of Neurology, Mayo Clinic, 200 1st Street SW, Rochester, MN 55905, USA
2
Department of Neurosurgery, Mayo Clinic, Rochester, MN, USA
etiology, intraventricular hemorrhage, midline shift, and anticoagulation were associated with increased mortality. On multivariate logistic regression, surgical intervention remained associated with decreased mortality (p < 0.0001; OR 0.20, 95 % CI 0.09–0.42). On subgroup analysis of the ICH cohort, surgical intervention was also associated with decreased mortality (p = 0.028). Conclusions Neurosurgical intervention for decompression in patients with trapped ventricle can have a measurable beneficial effect on early mortality. Keywords Hydrocephalus Trapping Ventricular Surgery Ventriculostomy Outcome Mortality
Introduction Trapped ventricle refers to focal or compartmentalized hydrocephalus usually caused by obstruction from a mass lesion. Choroid plexus secretion of cerebrospinal fluid (CSF) into an obstructed ventricle results in focal dilatation, mass effect, and neurologic decline. Trapping has been described at the lateral ventricle due to foramen of Monro obstruction, the temporal horn from trigone obstruction and the fourth ventricle (FV) from obstruction at the foramina of Luschka, Magendie, and the Aqueduct of Sylvius. Acute ventricular trapping may be caused by new lesions, suddenly expanding lesions, or as a result of sudden failure of CSF compensatory mechanisms in a patient with an existing mass. The etiology can be heterogeneous. Possible causes include congenital malformations, hemorrhage, trauma, tumors, and other various infectious and non-infectious diseases [1–10]. The existing literature on trapped ventricle mainly consists of case reports, small case series and neurosurgical technique discussions. Most of these studies are confined to cases of
123
Neurocrit Care
chronic trapped FV in the pediatric population, occurring as a late complication of shunting and successful lateral ventricular decompression for hydrocephalus secondary to meningitis or hemorrhage [11–13]. In adults, a commonly described site of trapping is the temporal horn of the lateral ventricle, but there were only 36 cases reported in the literature as of 2013 [14]. Therefore, very limited data exists on the epidemiology, treatments and outcomes of trapped ventricle from acute or worsening brain lesions with mass effect in the adult population. Likewise, the value of CSF diversion in the acute trapped ventricle has not been formally studied. Moreover, although trapped ventricle is well-known radiologically, there are no systematic studies on its clinical manifestations or on the effect of surgical decompression on patients’ outcomes. In this study, we analyzed a large cohort of adult patients with acutely trapped ventricle from different etiologies to characterize the prognosis and to assess the benefits of surgical intervention on early mortality.
Methods Following institutional IRB approval, a search algorithm was engineered to review all patient data from 2000 to 2012 in the Mayo Clinic electronic medical record (St. Fig. 1 CT scans (a–c) and MRI FLAIR (d) of patients with acute ventricular trapping caused by ICH (a), SDH (b), ischemic stroke (c), and GBM (d). (a) died without intervention, while (b), (c), and (d) survived after intervention.
123
Mary’s Hospital, Rochester, MN). Permutations of text patterns were used to search the records and identify cases potentially relevant to the study design. Six hundred and thirty consecutive patients were initially identified. MRI or CT findings were reviewed for each case and cases were only included if a trapped ventricle was diagnosed by a neuroradiologist or neurointensivist and confirmed by one of the investigators (GPE). When the diagnosis of trapped ventricle was complicated by ambiguous neuroimaging interpretation, a second investigator reviewed the case (AAR). If serial imaging did not clarify the diagnosis or the imaging remained indeterminate, these cases were excluded. Three hundred and ninety-two consecutive cases of trapped ventricle were confirmed and included in the analysis. Patient age, sex, medications, co-morbidities, etiology, site of entrapment, presence of intraventricular hemorrhage (IVH), degree of midline shift, Glasgow Coma Scale (GCS), presence of cerebral herniation with brainstem dysfunction (defined as coma with one or both pupils fixed and dilated, or absence of corneal, oculocephalic or oculovestibular reflexes), care withdrawal status, and patient outcome were recorded. Site of entrapment was categorized as right lateral ventricle (RLV), left lateral ventricle (LLV), third ventricle (TV), cerebral aqueduct, or FV. Combinations were also
Neurocrit Care Table 1 Group characteristics of trapped ventricle study cohort (n = 392) All groups (n = 392) n (%)
Non-intervention group (n = 171) n (%)
Intervention group (n = 221) n (%)
p
Male age mean, SD
58 (18)
67 (18)
57 (17)
ORIGINAL ARTICLE
Acutely Trapped Ventricle: Clinical Significance and Benefit from Surgical Decompression Gabriel L. Pagani-Este´vez1 • Philippe Couillard1 • Giuseppe Lanzino2 Eelco F. M. Wijdicks1 • Alejandro A. Rabinstein1
•
Ó Springer Science+Business Media New York 2015
Abstract Background Focal ventricular obstruction—trapped ventricle—results in cerebrospinal fluid accumulation, mass effect and possible clinical deterioration. There are no systematic studies on the benefit of surgical decompression in adults. Methods We reviewed patients admitted with acutely trapped ventricle on brain imaging to assess their prognosis and the effect of surgical intervention on 30-day mortality. Results Of the 392 patients with trapped ventricle, the most common causes were brain tumor (45 %), intracerebral hemorrhage (ICH) (20 %), and subdural hematoma (SDH) (14 %). Lateral ventricle trapping accounted for 97 % of cases. Two hundred and twenty-one patients (56 %) received a surgical intervention for trapped ventricle or its causes; 126 (83 %) were treated with craniotomy, 26 (17 %) with craniectomy, 30 (14 %) with external ventricular drain (EVD) alone, 23 (10 %) with ventriculoperitoneal shunt alone, and 16 (7 %) with endoscopic fenestration of the septum pellucidum. Surgical intervention was associated with mortality reduction from 95 % (n = 54) to 48 % (n = 11) in the ICH group, from 47 % (n = 27) to 12 % (n = 15) in the tumor group and from 90 % (n = 18) to 20 % (n = 7) in the SDH group (p < 0.001 for all comparisons). Univariate logistic analysis showed that surgical intervention and tumor etiology were associated with decreased mortality while age, ICH & Alejandro A. Rabinstein [email protected] 1
Department of Neurology, Mayo Clinic, 200 1st Street SW, Rochester, MN 55905, USA
2
Department of Neurosurgery, Mayo Clinic, Rochester, MN, USA
etiology, intraventricular hemorrhage, midline shift, and anticoagulation were associated with increased mortality. On multivariate logistic regression, surgical intervention remained associated with decreased mortality (p < 0.0001; OR 0.20, 95 % CI 0.09–0.42). On subgroup analysis of the ICH cohort, surgical intervention was also associated with decreased mortality (p = 0.028). Conclusions Neurosurgical intervention for decompression in patients with trapped ventricle can have a measurable beneficial effect on early mortality. Keywords Hydrocephalus Trapping Ventricular Surgery Ventriculostomy Outcome Mortality
Introduction Trapped ventricle refers to focal or compartmentalized hydrocephalus usually caused by obstruction from a mass lesion. Choroid plexus secretion of cerebrospinal fluid (CSF) into an obstructed ventricle results in focal dilatation, mass effect, and neurologic decline. Trapping has been described at the lateral ventricle due to foramen of Monro obstruction, the temporal horn from trigone obstruction and the fourth ventricle (FV) from obstruction at the foramina of Luschka, Magendie, and the Aqueduct of Sylvius. Acute ventricular trapping may be caused by new lesions, suddenly expanding lesions, or as a result of sudden failure of CSF compensatory mechanisms in a patient with an existing mass. The etiology can be heterogeneous. Possible causes include congenital malformations, hemorrhage, trauma, tumors, and other various infectious and non-infectious diseases [1–10]. The existing literature on trapped ventricle mainly consists of case reports, small case series and neurosurgical technique discussions. Most of these studies are confined to cases of
123
Neurocrit Care
chronic trapped FV in the pediatric population, occurring as a late complication of shunting and successful lateral ventricular decompression for hydrocephalus secondary to meningitis or hemorrhage [11–13]. In adults, a commonly described site of trapping is the temporal horn of the lateral ventricle, but there were only 36 cases reported in the literature as of 2013 [14]. Therefore, very limited data exists on the epidemiology, treatments and outcomes of trapped ventricle from acute or worsening brain lesions with mass effect in the adult population. Likewise, the value of CSF diversion in the acute trapped ventricle has not been formally studied. Moreover, although trapped ventricle is well-known radiologically, there are no systematic studies on its clinical manifestations or on the effect of surgical decompression on patients’ outcomes. In this study, we analyzed a large cohort of adult patients with acutely trapped ventricle from different etiologies to characterize the prognosis and to assess the benefits of surgical intervention on early mortality.
Methods Following institutional IRB approval, a search algorithm was engineered to review all patient data from 2000 to 2012 in the Mayo Clinic electronic medical record (St. Fig. 1 CT scans (a–c) and MRI FLAIR (d) of patients with acute ventricular trapping caused by ICH (a), SDH (b), ischemic stroke (c), and GBM (d). (a) died without intervention, while (b), (c), and (d) survived after intervention.
123
Mary’s Hospital, Rochester, MN). Permutations of text patterns were used to search the records and identify cases potentially relevant to the study design. Six hundred and thirty consecutive patients were initially identified. MRI or CT findings were reviewed for each case and cases were only included if a trapped ventricle was diagnosed by a neuroradiologist or neurointensivist and confirmed by one of the investigators (GPE). When the diagnosis of trapped ventricle was complicated by ambiguous neuroimaging interpretation, a second investigator reviewed the case (AAR). If serial imaging did not clarify the diagnosis or the imaging remained indeterminate, these cases were excluded. Three hundred and ninety-two consecutive cases of trapped ventricle were confirmed and included in the analysis. Patient age, sex, medications, co-morbidities, etiology, site of entrapment, presence of intraventricular hemorrhage (IVH), degree of midline shift, Glasgow Coma Scale (GCS), presence of cerebral herniation with brainstem dysfunction (defined as coma with one or both pupils fixed and dilated, or absence of corneal, oculocephalic or oculovestibular reflexes), care withdrawal status, and patient outcome were recorded. Site of entrapment was categorized as right lateral ventricle (RLV), left lateral ventricle (LLV), third ventricle (TV), cerebral aqueduct, or FV. Combinations were also
Neurocrit Care Table 1 Group characteristics of trapped ventricle study cohort (n = 392) All groups (n = 392) n (%)
Non-intervention group (n = 171) n (%)
Intervention group (n = 221) n (%)
p
Male age mean, SD
58 (18)
67 (18)
57 (17)
