https://doi.org/10.1007/s12028-017-0487-3 ORIGINAL ARTICLE Ventriculostomy and Risk of Upward Herniation in Patients with Obstructive Hydrocephalus from Posterior Fossa Mass Lesions Sherri A. Braksick 1,3 Benjamin T. Himes 2 Kendall Snyder 2 Jamie J. Van Gompel 2 Jennifer E. Fugate 1 Alejandro A. Rabinstein 1 Ó Springer Science+Business Media, LLC, part of Springer Nature 2017 Abstract Background Patients with posterior fossa lesions causing obstructive hydrocephalus present a unique clinical challenge, as relief of hydrocephalus can improve symptoms, but the perceived risk of upward herniation must also be weighed against the risk of worsening or continued hydrocephalus and its consequences. The aim of our study was to evaluate for clinically relevant upward herniation following external ventricular drainage (EVD) in patients with obstructive hydrocephalus due to posterior fossa lesions. Methods We performed a retrospective review of patients undergoing urgent/emergent EVD placement at our institution between 2007 and 2014, evaluating the radiographic and clinical changes following treatment of obstructive hydrocephalus. Results Even prior to EVD placement, radiographic upward herniation was present in 22 of 25 (88%) patients. The average Glasgow Coma Scale of patients before and after EVD placement was 10 and 11, respectively. Radiographic worsening of upward herniation occurred in two patients, and upward herniation in general persisted in 21 patients. Clinical worsening occurred in two patients (8%), though in all others the clinical examination remained stable (44%) or improved (48%) following EVD placement. Of the patients who had a worsening clinical exam, other variables likely also contributed to their decline, and cerebrospinal fluid diversion was likely not the main factor that prompted the clinical change. Conclusions Radiographic presence of upward herniation was often present prior to EVD placement. Clinically relevant upward herniation was rare, with only two patients worsening after the procedure, in the presence of other clinical confounders that likely contributed as well. Keywords External ventricular drain Upward herniation Herniation Posterior fossa Obstructive hydrocephalus Introduction Posterior fossa lesions, of any type, present a complex clinical challenge because the cranial vault does not allow accommodation for large space-occupying lesions. In subacute to chronic lesions (e.g., tumor) or in acute lesions & Sherri A. Braksick braksick.sherri@gmail.com 1 2 3 Division of Critical Care Neurology, Department of Neurology, Mayo Clinic, Rochester, MN, USA Department of Neurosurgery, Mayo Clinic, Rochester, MN, USA Department of Neurology, University of Kansas, 3901 Rainbow Blvd, Mail Stop 2012, Kansas City, KS 66160, USA (e.g., hemorrhage) there is a risk of both downward and upward herniation, through either the foramen magnum or the tentorial incisura, respectively, if differential pressures exist between either the supratentorial space or the cervicomedullary cistern relative to the posterior fossa. Urgent and sometimes emergent decompression is necessary to manage this life-threatening situation. In some instances, patients present with symptoms that are more consistent with obstructive hydrocephalus rather than direct parenchymal compression. In these cases, the question is whether a ventriculostomy with an external ventricular drain (EVD) for therapeutic cerebrospinal fluid (CSF) diversion can be safely performed. In patients who present with symptomatic obstructive hydrocephalus from lesions within the posterior fossa,
there is ongoing debate regarding the safety of CSF diversion due to the risk of developing or worsening preexistent upward herniation. The phenomenon of upward herniation has been previously demonstrated in pathologic series [1] as well as on computed tomography (CT) imaging studies [2] in patients who present with posterior fossa lesions. Dramatic clinical deterioration secondary to upward herniation is uncommon, but is reported [3]. Physiologically, there is concern for precipitating or worsening ascent of the midbrain through the tentorial incisura if CSF is diverted supratentorially due to the sudden change in CSF pressure dynamics between the supratentorial and infratentorial compartments. While the pathologic and radiographic findings are clear, the clinical implications and risks of CSF diversion in these patients have not been fully elucidated. There are reports of clinical worsening with findings consistent with midbrain compression following supratentorial CSF diversion [3 6], but these are few in number. Series have also shown that preoperative CSF diversion rarely causes symptoms of upward herniation and, if necessary, may be used as a bridge to definitive management of posterior fossa lesions, particularly in children [3, 5]. Among pediatric patients with posterior fossa tumors that undergo CSF diversion, a more common complication is post-procedure intratumoral hemorrhage, which can occur after drain or shunt placement [3, 4, 7, 8] and precipitate upward herniation from the increased mass effect within the posterior fossa. In adults, the clinical results of supratentorial CSF drainage in patients with obstructive hydrocephalus from posterior fossa mass lesions are even less defined, with anecdotal reports of both good outcome [9, 10] and postprocedure worsening [11]. The aims of our present study were to evaluate the safety of CSF diversion in adult and pediatric patients with mass lesions in the posterior fossa presenting with obstructive hydrocephalus, and attempt to identify risk factors for clinical worsening after EVD placement. Methods This study was approved by the Mayo Clinic Institutional Review Board. The medical records of patients who underwent EVD placement between 2007 and 2014 at Mayo Clinic, Rochester, MN, were reviewed for inclusion in this study. Patients who were identified as having urgent/ emergent placement of an EVD prior to surgical decompression for posterior fossa lesions causing acute neurological decline from obstructive hydrocephalus were included. Those who had communicating hydrocephalus and those undergoing elective EVD placement prior to decompressive surgery were excluded. Demographic information including age, gender, admitting diagnosis, and final pathologic diagnosis was collected. Examination findings, specifically the sum score of the Glasgow Coma Scale (GCS) [12], as documented by a neurologist or neurosurgeon, prior to and following EVD placement were recorded. The CT scans or magnetic resonance imaging (MRI) scans of the head obtained before and after EVD placement were personally reviewed by two of the study authors (JEF or AAR) to determine if upward herniation was present at either or both time points. Upward herniation was defined as effacement of the quadrigeminal cistern at the level of the tentorial incisura, as previously defined by Osborn et al. [2]. The clinical examination following EVD placement was compared to the pre-evd examination for evidence of improvement or worsening based on GCS sum scores. Statistical Analysis Descriptive summaries were reported as median and interquartile range (IQR) for continuous variables and as frequencies and percentages for categorical variables. Comparisons were made using the Fisher exact test. All the tests were 2-sided, and a p value \ 0.05 was considered significant. Analysis was performed using SAS version 9.4 (SAS Inc., Cary, NC). Results In total, 25 patients met inclusion criteria for this study. An additional 11 patients were evaluated but were excluded for various reasons, primarily due to the absence of a posterior fossa mass lesion (e.g., aneurysmal subarachnoid hemorrhage without a discrete mass lesion/obstruction) or the lack of appropriate peri-procedure imaging. The median age of our cohort was 46.4 years (IQR 22.5 64.5), and 48% (n = 12) were male. Five patients were 8 years of age or younger, and all others were older than 18. Additional patient characteristics are listed in Table 1. The median GCS before and after EVD placement was 10 (IQR 6 13.5) and 11 (IQR 8 15), respectively. The median GCS change per patient overall was 0, however. Despite this, nearly half of patients demonstrated at least a two-point improvement in their GCS following EVD placement. Only two patients (8% of the cohort) had clinical worsening after the procedure (see descriptions below). Twelve patients (48%) had a clinical improvement following EVD placement, with the median GCS
Table 1 Patient characteristic Patient number Age Gender Diagnosis GCS prior to EVD GCS following EVD Upward herniation on initial scan Upward herniation on follow-up imaging Required decompressive surgery 1 69 M Cerebellar hematoma 14 14 Yes Yes (stable) No 2 68 M SAH with IVH and 7 3 Yes Yes (improved) No cerebellar hematoma 3 54 M SAH and bilateral cerebellar hematomas 10 8 Yes Yes (stable) No 4 64 F Posterior fossa tumor 14 14 Yes Yes (stable) Yes 5 65 F Cerebellar hematoma 10 10 Yes Yes (stable) No 6 4 F Posterior fossa tumor 15 15 Yes Yes (stable) No 7 63 F Posterior fossa tumor 11 11 Yes Yes (stable) Yes 8 46 M Cerebellar hematoma 15 15 Yes Yes (stable) No 9 64 F Posterior fossa tumor 15 15 No No Yes 10 56 F Cerebellar ischemic 14 14 Yes Yes (stable) No stroke 11 19 M Trauma and cerebral 3 3 Yes Yes (worsened) No edema 12 56 F Postoperative edema 11 11 Yes Yes (stable) Yes 13 74 M Cerebellar ischemic 6 6 Yes Yes (stable) Yes stroke 14 64 F Posterior fossa tumor 6 9 Yes Yes (stable) Yes 15 75 M IVH 6 10 Yes No No 16 8 M Posterior fossa tumor 7 15 Yes Yes (improved) Yes 17 1 F Craniosynostosis 7 10 Yes Yes (stable) Yes 18 54 F Cerebellar hematoma 5 7 Yes Yes (worsened) No 19 44 M IVH 7 8 No No No 20 1 M Posterior fossa tumor 13 15 Yes Yes (stable) Yes 21 0 M Arachnoid cyst 3 8 Yes Yes (improved) No 22 53 F Posterior fossa tumor 13 15 Yes Yes (stable) No 23 61 F Cerebellar ischemic 13 15 Yes Yes (stable) Yes stroke 24 26 M Brainstem hemorrhage 12 13 No No No 25 72 F Cerebellar hematoma 3 9 Yes Yes (stable) No F female, GCS Glasgow Coma Scale, IVH intraventricular hemorrhage, M male, SAH subarachnoid hemorrhage improving from 7 (IQR 5.25 12.75) to 10 (IQR 8.25 15) within this group. The remaining 11 patients were clinically unchanged before and after the procedure. Prior to EVD placement, 22 patients (88%) were considered to have upward herniation by imaging criteria, including the two patients who had clinical worsening following EVD placement. After the procedure, imaging evidence of upward herniation resolved in one patient but remained present in the other 21. Within this group, herniation was felt to be improved (but persistent) in three, worse in two and unchanged in 16 patients following EVD placement. Among patients with radiographic herniation prior to intervention, the median GCS prior to EVD placement was 10 (IQR 6 13.25) and following EVD was 10.5 (IQR 8 15). See Fig. 1 for case examples. Three patients did not have herniation by radiographic criteria on their initial imaging. The median GCS prior to and after EVD were 12 (IQR 7 15) and 13 (IQR 8 15), respectively. Radiographically, the pre- and post-evd imaging was stable in all of these patients. In our cohort, radiographic upward herniation on pre-evd imaging was not associated with neurological decline (p = 1.0).
Neurocritical Care Fig. 1 Evidence of upward herniation before and after EVD placement. a CT scan in a 68-year-old man with significant posterior fossa hemorrhage who worsened clinically following EVD placement even though the degree of upward herniation was stable to improved on follow-up CT (b). CT scan (c) in a 54-year-old man with a spontaneous cerebellar hemorrhage worsened clinically following EVD placement in spite of no radiographic evidence of worsening upward herniation (d). Initial CT scan (e) in a 13-day-old boy presenting with acute hydrocephalus due to a posterior fossa arachnoid cyst. He demonstrated clinical improvement (GCS increase from 3 to 8) despite evidence of persistent upward herniation on follow-up MRI (f) Two patients had radiographic worsening of upward herniation after EVD placement. In both, there was evidence of upward herniation on the pre-evd scan, but the degree of herniation increased on the post-evd follow-up scan. One patient was clinically unchanged (GCS remained 3), while the other modestly improved (GCS increased from 5 to 7) following EVD placement. Cases of Clinical Deterioration After EVD Case 1 A 68-year-old male presented with a ruptured basilar artery aneurysm with intraventricular and cerebellar hemorrhages. The EVD was placed emergently on arrival to our
facility and within 12 h of symptom onset (the patient had been found down at home and exact time of onset was unclear). He also required endotracheal intubation. Following EVD placement and a conventional catheter angiogram, he had aneurysmal rebleeding with clinical worsening (GCS was initially 7 and decreased to 3). This deterioration happened in the context of refractory intracranial pressure elevation associated with EVD obstruction from blood clots. On repeat CT scan the upward herniation persisted, but was radiologically improved. Despite aggressive care, the patient s examination did not improve and he was eventually transitioned to palliative care after discussion with his family. Case 2 A 54-year-old male underwent an elective craniotomy for a suprasellar mass. Postoperatively his GCS was 10 and a CT of the head demonstrated bilateral cerebellar hematomas with diffuse subarachnoid blood, obstructive hydrocephalus and upward herniation. He underwent EVD placement within 2 h of the conclusion of his surgery. While on the CT scanner table for follow-up imaging he had a generalized seizure and was intubated before a post- EVD examination could be recorded. The CT was essentially unchanged with stable upward herniation. Once back in the ICU, his GCS was 8, although seizure, postictal state and the drugs used for intubation could have confounded that examination. After these initial complications, the patient steadily improved. He was extubated 4 days later, and his EVD was removed after 5 days. He made a full recovery after a course of acute inpatient rehabilitation. Discussion Despite very frequent evidence of radiographic upward herniation, placement of an EVD for obstructive hydrocephalus in patients with posterior fossa mass lesions was generally safe in our cohort. Two patients had clinical worsening after EVD placement (8% of all patients with obstructive hydrocephalus and 9% of those who also had radiographic upward herniation). However, in both cases there were other factors that may have contributed to the neurological deterioration despite, or at least in addition to, placement of the EVD. When CSF diversion can be safely established in patients with obstructive hydrocephalus, this modality may prevent the need for additional decompressive surgery in select patients, though others will require more definitive surgical management. Clinical improvement after drain placement was not uncommon in our patients, occurring in nearly half of our cohort. The safety of ventriculostomy in the great majority of our cases supports the emergent placement of an EVD for the treatment of symptomatic obstructive hydrocephalus from a posterior fossa mass lesion. Based on our experience, EVD placement should not be deferred due to a theoretical risk of clinical worsening from upward herniation, even if the radiographic findings appear concerning. As expected, there will always be exceptions, and one particular case of clinical decline following EVD placement at our institution prompted this current project. In that patient, an EVD was placed in the immediate postoperative setting following suboccipital craniectomy for a cerebellar metastatic lesion. Within 24 h his clinical condition improved and the EVD was removed. On postoperative day 2, due to somnolence and headache, CT was completed and demonstrated worsening hydrocephalus. Placement of an EVD resulted mild improvement initially; however, within minutes the patient progressed to coma with fixed pupils but retained respiratory effort. CT confirmed development of upward herniation. After the patient failed to improve, the decision was made to withdraw support. While this patient did not meet criteria to be included in our study as he had already undergone suboccipital decompression, this example demonstrates the potential danger of precipitating neurological decompensation in rare patients. These uncommon occurrences should be a reminder of the inherent risk of EVD placement in these situations, but should not preclude treatment when an EVD is otherwise indicated, as this patient already had undergone suboccipital decompression, but despite this developed symptomatic obstructive hydrocephalus. The reported cases of clinical worsening following CSF diversion, such as the case illustrated above, are scant, and uneventful or uncomplicated procedures are not often reported in the medical literature thus leading to a publication bias. Still, it is advisable to be particularly cautious when performing ventriculostomy in patients with posterior fossa mass lesions causing radiographic upward herniation. This intervention should be strictly reserved for patients with symptomatic and severe obstructive hydrocephalus in whom emergent evacuation or excision of the mass lesion or decompressive suboccipital craniectomy is either not immediately feasible or as a planned bridge to definitive treatment. As a safety measure, it is also reasonable to keep a relatively high level of drainage about the tragus to avoid sudden or excessive CSF outflow. There are limitations to our study. Our overall small numbers obtained at a large medical center emphasize the relative rarity of this clinical situation, but also precluded extensive statistical analysis with which to determine patient characteristics that may predispose to clinically significant upward herniation after CSF diversion, and limiting the current study to a descriptive analysis. We
were also fully dependent on the medical record due to the retrospective design of our study. It is possible that residual effects of sedating medications given during EVD placement falsely worsened some patients post-procedure GCS scores, resulting in a smaller perceived benefit of catheter placement in our cohort. A prospective analysis of these patients, as well as those with obstructive hydrocephalus but no mass lesion, with extensive data collection of patient variables for determination of factors that may predispose to clinical worsening would be ideal, but difficult due to the uncommon nature of this phenomenon. Conclusion Despite the common occurrence of upward herniation on radiographic imaging, urgent/emergent placement of an EVD was tolerated well by the great majority of patients with obstructive hydrocephalus from a posterior fossa mass lesion. Therefore, the presence of radiographic upward herniation should not preclude placement of an EVD in critically ill patients with symptomatic obstructive hydrocephalus. Compliance with Ethical Standards Conflict of interest The authors have no conflicts of interest to report. References 1. Cuneo RA, Caronna JJ, Pitts L, Townsend J, Winestock DP. Upward transtentorial herniation: seven cases and a literature review. Arch Neurol. 1979;36:618 23. 2. Osborn AG, Heaston DK, Wing SD. Diagnosis of ascending transtentorial herniation by cranial computed tomography. AJR Am J Roentgenol. 1978;130:755 60. 3. El-Gaidi MA, El-Nasr AH, Eissa EM. Infratentorial complications following preresection CSF diversion in children with posterior fossa tumors. J Neurosurg Pediatr. 2015;15:4 11. 4. Epstein F, Murali R. Pediatric posterior fossa tumors: hazards of the preoperative shunt. Neurosurgery. 1978;3:348 50. 5. Raimondi AJ, Tomita T. Hydrocephalus and infratentorial tumors: incidence, clinical picture, and treatment. J Neurosurg. 1981;55:174 82. 6. Yadav G, Sisodia R, Khuba S, Mishra L. Anesthetic management of a case of transtentorial upward herniation: an uncommon emergency situation. J Anaesthesiol Clin Pharmacol. 2012;28:413 5. 7. Vaquero J, Cabezudo JM, de Sola RG, Nombela L. Intratumoral hemorrhage in posterior fossa tumors after ventricular drainage. Report of two cases. J Neurosurg. 1981;54:406 8. 8. Elgamal EA, Richards PG, Patel UJ. Fatal haemorrhage in medulloblastoma following ventricular drainage: case report and review of the literature. Pediatr Neurosurg. 2006;42:45 8. 9. Seelig JM, Selhorst JB, Young HF, Lipper M. Ventriculostomy for hydrocephalus in cerebellar hemorrhage. Neurology. 1981;31:1537 40. 10. Adamson DC, Dimitrov DF, Bronec PR. Upward transtentorial herniation, hydrocephalus, and cerebellar edema in hypertensive encephalopathy. Neurologist. 2005;11:171 5. 11. McDougall CM, Jack A, Raymond J, Bojanowski MW, Darsaut TE. Angiographic demonstration of upward transtentorial herniation. Can J Neurol Sci. 2014;41:82 3. 12. Teasdale G, Jennett B. Assessment of coma and impaired consciousness: a practical scale. Lancet. 1974;2:81 4.