Case Report
Dillon Burks, MD. Pediatric Critical Care Fellow. Department of Pediatrics. Division of Critical Care Medicine. University of Texas Health Science Center at Houston, Houston, TX, USA.
Beth Anne Cavanaugh, MD. Assistant Professor. Department of Pediatrics. Division of Pediatric Neurology. The University of Tennessee Health Science Center. Memphis, TN, USA.
Fiza Laheji, MD. Vascular Neurology Fellow. Department of Pediatrics. Division of Pediatric Neurology. The University of Texas Southwestern Medical Center. Dallas, TX, USA.
Wilmot Bonnet, MD. Assistant Professor. Department of Pediatrics and Neurology. Division of Pediatric Neurology. The University of Texas Southwestern Medical Center. Dallas, TX, USA.
Michael M Dowling, MD, PhD. Professor. Department of Pediatrics and Neurology. Division of Pediatric Neurology. The University of Texas Southwestern Medical Center. Dallas, TX, USA.
Lisa Pabst, MD. Assistant Professor. Department of Pediatrics. Division of Neurology. University of Utah, Primary Children’s Hospital. Salt Lake City, UT, USA.
Stuart Fraser, MD. Assistant Professor. Department of Pediatrics. Division of Child and Adolescent Neurology. The University of Texas McGovern Medical School. Houston, TX, USA.
Corresponding author
Stuart Fraser
Mailing Address: 6410 Fannin St. Suite 1535 Houston, TX 77030
Email Address: stuart.m.fraser@uth.tmc.edu
Telephone: (713) 500-7164
Fax: (713) 500-0719
Copyright © belongs to author(s)
All rights reserved.
Any redistribution or reproduction of part or all of the contents in any form is prohibited other than the following:
- you may print or download to a local hard disk extracts for your personal and non-commercial use only
- you may copy the content to individual third parties for their personal use, but only if you acknowledge the website as the source of the material
You may not, except with our express written permission, distribute or commercially exploit the content. Nor may you transmit it or store it in any other website or other form of electronic retrieval system.
Abstract
Childhood-onset stroke is a morbid condition affecting approximately 1 in 25,000 children annually. Increased intracranial pressure is a known cause of stroke. In this case series of eight patients, we describe different patterns of cerebral infarction after acutely raised intracranial pressure, including posterior cerebral, watershed, and cerebellar infarctions. Medical providers should be aware of these patterns and their association with increased intracranial pressure.
Introduction
Pediatric stroke is a morbid condition with an incidence of about 1 in 25,000 children annually.1 Previously under-recognized and under-reported, pediatric stroke has gained more exposure with the advancement of imaging techniques and organized and collaborative research efforts like the International Pediatric Stroke Study.2,3 Posterior circulation stroke, in particular, is prone to high rates of recurrence in children, and there is much to learn regarding posterior circulation stroke pathophysiology and mechanisms.4,5
Pediatric patients experiencing posterior circulation stroke have a median age of 7 to 8 years, are mostly male, and often have no past medical history.1,6 Signs and symptoms upon presentation may include dysarthria, visual deficits, ataxia, vertigo, hemiparesis, nausea and vomiting. Vertebral artery dissection due to recent trauma remains the most commonly reported cause (25-50%), followed by arteriopathies.6 However, transtentorial herniation has long been posited to also be a potential cause of posterior circulation stroke in children and adults through impingement of the posterior cerebral artery on the tentorium leaflet.7 Acute hydrocephalus has been reported to be associated with posterior cerebral artery infarction as well.8,9 The incidence and frequency of the phenomenon of acute hydrocephalus causing cerebral infarction in children is not reported in the literature to our knowledge. While there are case reports of acute infarction in the setting of hydrocephalus, there are few case series addressing different patterns that may be seen in this phenomenon. In this paper, we present a series of patients presenting with instances of increased intracerebral pressure and hydrocephalus of varying causes who developed a variety of different patterns of cerebral infarction. We seek to illustrate some of the patterns of strokes that can present in raised intracranial pressure and hydrocephalus, demonstrate the heterogeneity of stroke in these conditions, and review mechanisms of stroke in these conditions including arterial compression, arterial entrapment, and watershed infarction.
Case Presentations
Case 1: A 3-week-old boy born at 35 weeks gestational age presented to the Emergency Department for 3 days of lethargy and 1 day of hypothermia. They were subsequently found to be hyponatremic and hypoglycemic, and cerebrospinal fluid (CSF) studies were consistent with bacterial meningitis. Blood and CSF cultures grew group B streptococcus. Initial computed tomography (CT) showed severe ventriculomegaly and hydrocephalus without midline shift, and the patient underwent emergent external ventricular drain (EVD) placement following admission. A follow-up magnetic resonance image (MRI) obtained 72 hours later showed restricted diffusion in the right occipital lobe consistent with right posterior cerebral artery infarction (Figure 1) as well as ventriculitis. Follow up CT angiogram of the brain and neck did not demonstrate any vasculopathy or other vascular abnormalities and echocardiogram demonstrated normal cardiac function and anatomy. Coagulation studies were normal during the hospitalization.
(View Figure 1 in the attached PDF): Figure 1: Axial diffusion weighted magnetic resonance imaging demonstrating restricted diffusion in the right occipital lobe.
Case 2: A 2-year-old boy with a history of achondroplasia presented to the emergency department after 5 days of increasing lethargy, emesis, and diarrhea. He had a reported history of tethered cord status post release and hydrocephalus s/p placement of lumboperitoneal shunt sometime in his first year of life. Notably, the patient’s prior procedures were done in a separate country and no medical records were available for review. The patient worsened clinically one day after admission, developing seizures and apnea. He had no recorded periods of hypotension. MRI obtained at that time revealed changes consistent with ischemia in the watershed areas of the left middle cerebral artery (MCA), left posterior cerebral artery (PCA), right anterior cerebral artery (ACA), right MCA, right PCA, and the cerebellum (Figure 2). He was taken for external ventricular drain placement emergently. Initial intracranial pressure measurement with the EVD was 28 mm Hg, which decreased to 5-10 mm Hg in the following days. The patient underwent right ventriculoperitoneal shunt placement once clinically stable. Magnetic Resonance Angiogram (MRA) of the brain and neck were unremarkable and demonstrated no asymmetry in the intracranial arteries. Echocardiogram and hypercoagulable evaluation did not reveal any abnormalities.
(View Figure 2 in the attached PDF): Figure 2: Diffusion weighted magnetic resonance imaging demonstrating restricted diffusion in the left greater than right watershed distribution (A) and the cerebellum bilaterally (B).
Case 3: An 11-year-old girl with a history of ventriculoperitoneal shunt placement for obstructive hydrocephalus presented with 1.5 days of lethargy and headache. CT of the brain in the ER demonstrated increased ventricular size compared to previous, with marked hydrocephalus with relative sparing of the fourth ventricle and mild transtentorial herniation. She was diagnosed with shunt malfunction. She underwent an emergent VP shunt revision, and no follow up neuroimaging was obtained at the time. She was discharged home on post operative day number 3. She then returned to the ER on post operative day number 9 with 3 days of somnolence, decreased oral intake, and decreased urine output. MRI of the brain with diffusion weighted magnetic resonance imaging (DW-MRI) revealed an isolated right occipital infarction which was not appreciated on her CT of the brain 9 days earlier (Figure 3). MRA of the head and neck was normal at the time.
(View Figure 3 in the attached PDF): Figure 3: Diffusion weighted magnetic resonance imaging demonstrating restricted diffusion in the right occipital lobe.
Case 4: An 11-year-old girl presented with seizures and coma. She was intubated in the field and brought to the ER. CT head obtained on arrival demonstrated spontaneous intraparenchymal hemorrhage in the left thalamus with intraventricular extension. She had hydrocephalus and clinical signs of increased intracranial pressure on presentation, with an opening pressure of 20cm of water on initial external ventricular drain placement in the ER. CT Angiogram of the brain and neck demonstrated a left posterior cerebral artery aneurysm, with no other vessel abnormalities found. T2 fluid-attenuated inversion recovery (FLAIR) imaging taken the day after admission demonstrated a left occipital infarction (Figure 4). Echocardiogram demonstrated normal anatomy and the presence of intracardiac shunting of bubbles.
(View Figure 4 in the attached PDF): Figure 4: T2 FLAIR magnetic resonance imaging demonstrates increased T2 signal in the left occipital lobe.
Case 5: A 16-year-old boy with well-controlled focal epilepsy as well as 26 week preterm birth complicated by intraventricular hemorrhage and hydrocephalus requiring a ventriculoperitoneal shunt presented with 3 days of worsening lethargy and vomiting. He was emergently intubated and found to have severe hydrocephalus on initial CT scan of the head with fracture of the distal shunt catheter. There were no signs of midline shift or herniation, however. He was diagnosed with ventriculoperitoneal shunt malfunction and hydrocephalus. He was taken for emergent shunt revision. Repeat head CT 36 hours afterwards demonstrated decompressed ventricular system but there was a new hypodensity in the left occipital lobe and left anterior inferior cerebellum. Diffusion weighted magnetic resonance imaging on hospital day 4 demonstrated infarction in the right and left occipital lobes, brain stem, and left cerebellum (Figure 5). Time of flight magnetic resonance angiography of the head and neck and transthoracic echocardiogram with bubble study did not demonstrate any abnormalities.
(View Figure 5 in the attached PDF): Figure 5: Diffusion weighted magnetic resonance imaging demonstrates infarction in (A) the right occipital lobe, (B) the brain stem and (C) the left cerebellum.
Case 6: A 13-year-old girl with Chiari II malformation and shunted hydrocephalus presented with 1 day of progressive headache followed by unresponsiveness. CT of the brain revealed enlarged ventricles and she was diagnosed with shunt failure and went to the operating room for emergent shunt revision. Her post-operative course was complicated with new focal neurologic deficit noted upon waking from anesthesia about 8 hours after surgery, including right facial numbness, right facial droop, dysphagia, multidirectional nystagmus, and bilateral lower extremity weakness. Subsequent MRI demonstrated bilateral posterior inferior cerebellar artery (PICA) infarctions (Figure 6). Echocardiogram and hypercoagulable workup were not obtained, but CT angiogram of the brain and neck performed immediately after the MRI brain was normal.
(View Figure 6 in the attached PDF): Figure 6: T2 FLAIR magnetic resonance imaging demonstrates subtle changes in the bilateral inferior cerebellar hemispheres.
Case 7: A ten year old boy with sickle cell disease (hemoglobin SC), attention deficit hyperactivity disorder, autism, and epilepsy was admitted to the hospital with a vaso-occlusive crisis with leg pain as well as hypoxemic acute respiratory failure in the setting of acute chest syndrome. He had a complicated hospital course, including a pulmonary embolus with a deep venous thrombosis of the leg and suspected endocarditis with a thrombus versus vegetation noted on echocardiogram. He was anticoagulated for this reason. His hospital stay was further complicated by a large right subdural hematoma (Figure 7A) which required emergent evacuation and a craniotomy. CT of the brain did not demonstrate an infarction. An MRI of the brain obtained 3 days after the surgery demonstrated a right occipitoparietal infarction (Figure7B). MRA of the head and neck did not demonstrate abnormality in the head or neck vessels. Echocardiogram with bubble study was obtained after the stroke which was negative for agitated bubble study. Hemoglobin S level ranged between 6.3-7.8 at the time.
(View Figure 7 in the attached PDF): Figure 7: Computed tomography image of the brain demonstrates right subdural hemorrhage with midline shift and transtentorial herniation (A). Follow up MRI of the brain on post operative day 3 demonstrates diffusion restriction in the right occipital lobe, consistent with infarction(B).
Case 8: A 14 month old girl with a history of 23-week preterm birth, post hemorrhagic hydrocephalus requiring a ventriculo-atrial shunt, epilepsy, and short gut syndrome presented to the ER with 4 days of lethargy and vomiting. A CT of the brain was obtained and demonstrated interval increase in the size of the lateral and 3rd ventricles suggestive of raised intraventricular pressure (Figure 8A). She was diagnosed with shunt malfunction and underwent shunt revision emergently. MRI obtained the following day did not show any areas of restricted diffusion. However, repeat MRI 4 days after presentation was significant for acute infarction in the in the right posterior mesial temporal lobe and the inferior gyri of the right temporal and occipital lobes in the distribution of the posterior cerebral artery (Figure 8B,C). Vessel imaging was not obtained at the time due to the hypothesis that posterior circulation infarct was likely secondary to PCA compression. Echo was not obtained and hypercoagulable work up was not completed. The patient had a follow up MRI 4 months later which showed markedly diminished size of ventricles and no new infarcts.
(View Figure 8 in the attached PDF): Figure 8: CT of the brain at presentation was significant for interval increase in the size of the lateral ventricles, consistent with hydrocephalus, with effacement of the extra-axial spaces. There was no midline shift(A). Follow up diffusion weighted magnetic resonance images demonstrated acute infarction in the right posterior mesial temporal lobe and the inferior gyri of the right temporal and occipital lobes in the distribution of the posterior cerebral artery (B, C).
(View Table 1 in the attached PDF): Table 1: Summary of proposed mechanisms of PCA infarctions and cases that experience them.
Discussion
In this paper, we demonstrate eight cases with cerebral infarcts in posterior circulation and/or watershed distributions in the setting of suddenly raised intracranial pressure. While this association has never been formally investigated in animal models to our knowledge, the anatomic course of the posterior cerebral artery warrants increased attention to possible ischemic sequelae following episodes of increased intracranial pressure due to the presence of the rigid tentorium cerebelli.
The Monro-Kellie doctrine warrants a brief review in this discussion, as expansive tumors, empyemas, hydrocephalus, and other space-occupying pathology are often detractors from the cerebral blood flow and contributors to the local ischemia seen in patients who experience stroke. The Monro-Kellie doctrine states that, conceptually, there is a fixed volume available within the cranial vault. A complex physiologic balancing act is then performed by the cranial vault’s three actors: the CSF (10% of the volume), the blood flow (10% of the volume), and the brain parenchyma (80% of the volume). Any interruptions to this phenomenon come at the cost of the actors. This may take the form of decreased available CSF, reduced cerebral blood flow, or diminished space for the brain parenchyma. A tumor, for example, may grow to occupy an arbitrary 5% of the total available volume. This volume is then subtracted from available volume of the CSF, blood flow, or parenchyma. In our proposed mechanism of PCA infarction, periods of acute hydrocephalus may cause disruptions of cerebral blood flow but are also aided by the rigidity of the tentorium cerebelli as the PCA becomes compressed between the tentorium’s edge and the parenchyma in settings of increased ICP and perturbations of the Monro-Kellie doctrine.
The cerebral perfusion pressure (CPP) is the driving pressure that dictates blood flow and oxygen delivery to the brain. It is calculated by the following equation, where CPP = cerebral perfusion pressure, MAP = mean arterial pressure, and ICP = intracranial pressure:
CPP=MAP-ICP
Preserving cerebral perfusion pressure is one of the primary goals of every clinician following any type of neurologic injury or pathology to prevent the adverse effects of ischemia that a compromised neurologic vasculature may lead to. Elevated ICP without a compensatory rise in the MAP may jeopardize the cerebral perfusion pressure and its dependent brain parenchyma as well as predispose the vasculature to external forces such as compression, underscoring why preserving this value is of such a large concern in the acute phase following neurologic insult.
Occipital lobe infarctions due to ipsilateral herniation across the tentorium have previously been described as well as contralateral infarction following instances of mass effect and midline shift.7,10 Both pathophysiological outcomes are thought to be attributed to the inflexibility of the tentorium cerebelli and the anatomical path the posterior cerebral artery traverses as it climbs the midbrain and enters the occipital lobe (Figure 9). Both clinical outcomes result from the presence of increased intracranial pressure, though the hypothetical mechanisms between the two may slightly differ. While ipsilateral infarctions can result from the downward vector applied to the tentorium with increasing volume (such as that of hemorrhage, hydrocephalus, malignancy, cyst formation, or other space-occupying lesions), contralateral infarctions are thought to be the product of midline shift and the lateral vector applied to the upper brainstem. This has been hypothesized to set off a chain reaction of events termed the “accordion effect”, first distorting the interventricular foramen as midline deviation worsens, then obstructing the outward flow of the lateral ventricle.7,10 This causes ventriculomegaly of the affected ventricle, decreases the available intracranial volume, and applies a downward vector to the tentorium, potentially impinging upon the ipsilateral posterior cerebral artery, as in the previous example of ipsilateral occipital lobe infarctions. Notably, the thalamus is typically supplied by perforating branches of the basilar artery and posterior cerebral artery that segment off prior to the tentorial membrane. This is consistent with the lack of thalamic infarction seen in cases 3, 4, and 5. Additionally, individual variations in collateral flow, metabolic demands related to seizures, and intracranial pressure distribution may explain the variations in PCA infarction locations seen in cases 1, 3, 4, and 5.
Bilateral posterior cerebral artery infarctions have also been reported in an adult patient with tuberous sclerosis complex and the presence of subependymal giant cell astrocytomas as well as patients with shunt malfunction presenting with hydrocephalus.8,10 The bilateral distribution of flow restriction was posited to be caused by brief instances of increased intracranial pressure associated with seizures and the “accordion effect” briefly interrupting the flow in the bilateral posterior cerebral arteries. This mechanism also likely plays a central role in these patients, with an unknown distribution of vectors and varying instances of increased intracranial pressure associated with each of their underlying comorbidities. These presentations consist of acutely elevated intracranial pressure and, likely, a continued downward vector of force applied across the plane of the tentorium cerebelli. Due to the variability of the cases, these episodes of increased intracranial pressure may result in only watershed infarcts, only isolated PCA infarctions, or any combination involving the two. Moreover, some episodes of restricted flow may be more transient than others, resulting in an asymmetric distribution of the ischemia when evaluated with imaging.
Another possible explanation for PCA infarction, particularly in cases associated with underlying infection or meningitis such as in case 1, is focal cerebral arteriopathy (FCA).7,11,12 Fink et al reported this etiology as the most prominent cause of PCA stroke in a group of 43 children in Switzerland (25%). The primary defect underlying arteriopathy is luminal stenosis of the affected vessels, such as found in inflammatory disease processes or vasculitis. Bacterial meningitis may precipitate this arteriopathy, given its systemic inflammatory effects and changes in circulating cellular components. 8,10
It should also be noted that interruptions to the cerebral blood flow need not be uniform across bilateral vessels. Many of the cases presented here demonstrate R>L or L>R ischemic effects and the presence of seizures around the time that the ischemia was noted on imaging. This may be multifactorial in cause. The most straightforward explanation is that the hydrocephalus caused an increase in ICP and negatively impacted cerebral blood flow, causing ischemia. It is also a possibility that in the presence of increased ICP, a focal seizure could have increased metabolic demand and blood flow to that area, shunting away from and further under-perfusing the contralateral region and resulting in the ischemia we see on the images. The increase in blood flow while experiencing a seizure may have also worsened the ICP, in turning causing more shifting of the parenchyma, more mass effect, and ultimately more restriction to blood flow.
In this case series, pediatric patients who developed acute hydrocephalus from varying causes experienced increased intracranial pressure. Subsequently, these patients were found to have imaging findings consistent with restricted flow and ischemia in the posterior cerebral artery distributions and/or watershed distributions of the anterior cerebral artery/middle cerebral artery and middle cerebral artery/posterior cerebral artery. While decreased cerebral perfusion pressure can easily explain watershed infarction, the mechanism of these posterior cerebral infarctions is not as well understood. The mechanisms may be similar (if not identical) to the interplay between mass effect, midline deviation, disruption of the Monro-Kellie doctrine, and vectors of force applied across the tentorium cerebelli, as described in the above isolated posterior cerebral infarctions. The literature is sparse with investigations of pre-stroke hydrocephalus. Rather, recent literature focuses on the occurrence of post-stroke or hemorrhage development of hydrocephalus as a complication instead. 8,10 Analogous to pressure of the tentorium cerebelli against the posterior cerebral artery, it is possible that displacement of the brainstem and cerebellum downward against the relatively fixed position of the vertebrobasilar arteries may lead to reduced perfusion in penetrating arteries as can be seen in the median and paramedian artery territory injury in Figure 5B. In this way, a transient “kinking” of penetrating arteries may cause hypoperfusion, which is alleviated after the brain returns to its normal position. A similar mechanism may happen at the cerebellar arteries and their penetrating branches.
This small case series has several weaknesses that should be addressed. 8,10Some additional details that were not available from chart review would be helpful in better characterizing the underlying pathophysiology associated with each stroke. Intracranial pressure monitoring through devices such as EVD would be useful, especially in distinguishing between periods of extra-arterial compression vs. arteriopathy, but were not available in all cases. The number of hours from the onset of presenting symptomology to the onset of stroke-like clinical symptoms, which was not clear in all cases, would also be useful as more acute presentations may weaken the evidence for arteriopathy and inflammatory processes. Other organ-specific studies and imaging, such as echocardiogram to rule out some of the mechanisms of thromboembolism, would also be helpful. In future studies, obtaining MRA at the earliest signs of suspected stroke or conducting serial transcranial Doppler studies may help to better delineate between periods of extra-arterial compression from increased ICP or mass effect, transient arteriopathy, or other modalities of stroke. Future research that allows us to better understand the precise mechanisms of posterior cerebral artery strokes in the setting of increased intracranial pressure, such as the role of endothelial injury, could provide insights on strategies for secondary stroke prevention.
While not formally reported, the presentation of these patients, coupled with the previous scattered reports of posterior cerebral artery infarction following hemorrhage or increased seizure-frequency, should serve as a cautionary tale and highlight the susceptibility of the posterior circulation and occipital lobe to instances of restricted flow during periods of increased intracranial pressure and metabolic demands. Recognition of the increased danger of ischemia, clinical presentation, and correction of the underlying cause remain the best identifiers and treatment for this unfortunate outcome.
Author Contributions:
Dr. Dillon Burks drafted the manuscript abstract, introduction, and discussion.
Dr. Cavanaugh contributed a case and magnetic resonance imaging.
Dr. Laheji provided cases and case descriptions.
Dr. Bonnet provided cases and case descriptions.
Dr. Dowling provided cases and magnetic resonance imaging. He edited the manuscript for content.
Dr. Pabst contributed a case and magnetic resonance imaging.
Dr. Fraser contributed two cases and magnetic resonance imaging. He edited the manuscript for content and formatted the manuscript for submission.
Conflict of Interest:
The authors report no relevant financial conflicts of interest for this work. MM Dowling recused himself from editorial decision making on this manuscript. Editorial decisions were made by the Associate Editors for this manuscript.
References
- Ferriero DM, Fullerton HJ, Bernard TJ, et al. AHA / ASA Scientific Statement Management of Stroke in Neonates and Children. 2019.
- Mackay MT, Wiznitzer M, Benedict SL, Lee KJ, deVeber GA, Ganesan V. Arterial ischemic stroke risk factors: The international pediatric stroke study. Ann Neurol 2011;69(1):130–40.
- Sun LR, Linds A, Sharma M, et al. Cancer and Tumor-Associated Childhood Stroke: Results From the International Pediatric Stroke Study. Pediatr Neurol 2020;111:59–65.
- Braga BP, Sillero R, Pereira RM, et al. Dynamic compression in vertebral artery dissection in children: apropos of a new protocol. Child’s Nervous System 2021;37(4):1285–93.
- Fullerton HJ, Wintermark M, Hills NK, et al. Risk of Recurrent Arterial Ischemic Stroke in Childhood. Stroke 2016;47(1):53–9.
- Fink M, Slavova N, Grunt S, et al. Posterior Arterial Ischemic Stroke in Childhood. Stroke 2019;50(9):2329–35.
- Ansari M, Chang GY. Pathophysiology of Stroke in the Contralateral Posterior Cerebral Artery Distribution from a Tentorial Herniation. Journal of Clinical Neurology 2017;13(1):101.
- Gülşen İ. Şant Disfonksiyonuna Bağlı sekonder oksipital İnfarkt: Olgu Sunumu. Journal of Contemporary Medicine 2016;6(0).
- Carey S, Rafay Mubeen. Pediatric Posterior Arterial Ischemic Stroke: A Summary of Current Knowledge. Pediatric Stroke Journal 2021;2:1–38.
- Kim J-G, Choi JC, Kim HJ, et al. Bilateral posterior cerebral artery stroke following transtentorial herniation caused by a subependymal giant cell astrocytoma in a patient with tuberous sclerosis: a case report. Journal of Neurocritical Care 2021;14(2):98–102.
Patterns of Cerebral Infarction
