Posterior Reversible Encephalopathy Syndrome Arising from Simultaneous Cranioplasty and Ventriculoperitoneal Shunting: A Case Report

Kensuke IKEDA; Keisuke MARUYAMA; Hiroki KAGIWATA; Atsushi YAMAMICHI; Kei OKADA; Shoko FUJII; Kuniaki SAITO; Hirofumi NAKATOMI; Teruyuki HIRANO; Motoo NAGANE

doi:10.2176/jns-nmc.2025-0067

Abstract

Posterior reversible encephalopathy syndrome is a well-known condition that causes reversible vasogenic edema, mainly in the occipital lobe. However, no guideline for its diagnosis or treatment has been established to date. While many atypical cases have been reported in recent years, posterior reversible encephalopathy syndrome associated with cranioplasty has not yet been reported. A man in his 50s underwent right decompressive craniectomy for brain swelling secondary to ischemic stroke. He was transferred to another hospital for rehabilitation 9 months later and was again admitted for cranioplasty after 2 months. Because of sustained brain distension, right cranioplasty was performed simultaneously with ventriculoperitoneal shunting. No sudden change in blood pressure occurred during the perioperative period. However, generalized convulsions occurred postoperatively. Magnetic resonance imaging showed signal changes in the right frontal lobe, left basal ganglia, left thalamus, and right occipital lobe. When cranioplasty was performed simultaneously with ventriculoperitoneal shunting for a skull defect complicated by brain distension, intraoperative cerebrospinal fluid drainage and continuous cerebrospinal fluid drainage by shunts may have caused a sudden decrease in intracranial pressure and an increase in cerebral perfusion pressure, possibly leading to the development of posterior reversible encephalopathy syndrome.

Introduction

Posterior reversible encephalopathy syndrome (PRES) is a syndrome that causes reversible vasogenic edema, mainly in the occipital lobe, resulting in non-specific symptoms, such as headache, impaired consciousness, seizures, and visual impairment. The main cause is believed to be interstitial edema due to the breakdown of the blood-brain barrier following a sudden rise in blood pressure, which causes cerebral perfusion to exceed the limits of autoregulatory capacity.¹^,²⁾ Since there are only a few sympathetic nerves in the territory of the posterior cerebral artery, they are considered potentially vulnerable to overperfusion. Its symptoms and imaging results commonly improve once blood pressure stabilizes and cerebral perfusion normalizes.³⁾

Although various PRES conditions have been reported, their pathogenesis has not yet been established, and there is no guideline for their diagnosis or treatment. Previous literature on the development of PRES in association with cranioplasty is not yet available. Here, we report a case of postoperative PRES associated with simultaneous cranioplasty and ventriculoperitoneal (VP) shunting in a patient with extensive cerebral infarction who underwent decompressive craniectomy.

Case Report

A man in his 50s was brought to our hospital with left hemiparesis and conjugate deviation of the eyes to the right side. Computed tomography (CT) and magnetic resonance (MR) imaging showed acute cerebral infarction in the territory of the right anterior cerebral artery and right middle cerebral artery. However, intravenous fibrinolysis or trans-arterial revascularization therapy was not indicated since the infarction was extensive. Anti-edema medications such as mannitol and concentrated glycerin-fructose were administered, but the acute cerebral edema worsened the day after admission. Therefore, decompressive craniectomy was performed. The postoperative course was favorable, and there were no complications. After 9 months, the patient was transferred to another hospital for rehabilitation. The modified Rankin Scale (mRS) score was 4.

Two months later, the patient was readmitted for cranioplasty. On admission, the Glasgow Coma Scale was E4V5M6, and there was no aphasia or hemispatial neglect. However, motor palsy and sensory deficits were observed in the left extremities. The patient had a skull defect in the right frontotemporal region that was distended without tension. CT showed low-density areas in the territory of the right anterior and right middle cerebral arteries, and the artificial dura mater was distended beyond the skull margin. On MR imaging, the infarcted foci were mostly cystic. MR angiography showed occlusion of the right internal carotid artery. The right middle cerebral artery was marginally depicted by cross flow through the anterior and posterior communicating arteries (Fig. 1A-D).

Fig. 1

Preoperative CT and MR imaging.

A: CT showed low-density areas in the territory of the right anterior cerebral artery and right middle cerebral artery, and the artificial dura mater was distended beyond the skull margin. B: Diffusion-weighted MR imaging showed a low signal in the territory of the right anterior cerebral artery and right middle cerebral artery. C: Fluid attenuated inversion recovery imaging showed that the infarcted area was cystic, reflecting necrosis and malacia of brain tissue, and contained abundant liquid components inside. D: MR angiography showed occlusion of the right internal carotid artery. The right middle cerebral artery was marginally depicted by cross flow through the anterior communicating artery and the posterior communicating artery.

CT: computed tomography; MR: magnetic resonance

Cranioplasty was performed in conjunction with a VP shunting since the infarcted foci were distended and complicated by brain distension. The patient's head was rotated to the left. A ventricular catheter was placed in the right posterior horn of the lateral ventricle, and 50 ml of spinal fluid was drained in 1 min. Next, a skin incision made during the initial surgery was made again. Preoperative CT showed that the artificial dura was distended beyond the bony margin, but intraoperative findings showed that the artificial dura was sunken rather than distended due to preceded VP shunting. Removal of the artificial dura exposed the sunken cerebral surface. The dural plasty was performed using the fascia lata. The artificial bone was replaced in the cranial defect, and the wound was closed. The intraoperative systolic blood pressure was 90-120 mmHg, with no excessive blood pressure changes.

Shortly after surgery and the cessation of general anesthesia, generalized tonic seizures were observed. Diazepam was administered, and the seizure was temporarily aborted. However, the patient did not wake up, and the seizure reappeared 30 mins later. The patient was diagnosed to be in status epilepticus, was intubated, and placed on a ventilator. The patient was sedated with continuous propofol, and levetiracetam was administered. The right frontal lobe, left basal ganglia, left thalamus, and right occipital lobe showed a high signal on diffusion-weighted MR imaging and fluid-attenuated inversion recovery imaging. The right frontal lobe and right occipital lobe showed a high signal, the left basal ganglia and left thalamus showed a low signal on the apparent diffusion coefficient map (Figs. 2 and 3). MR angiography revealed no new stenosis or occlusion of the anterior, middle, or posterior cerebral arteries (Fig. 4A). MR venography revealed no defect in cerebral venous sinuses (Fig. 4B and C). These imaging findings were not consistent with those of cerebrovascular disorders, such as acute cerebral infarction or cerebral venous sinus thrombosis. Continuous electroencephalography revealed periodic hemispheric discharges in the left frontal, temporal, and occipital regions. We speculated that the intracranial pressure in the patient decreased rapidly due to intraoperative cerebrospinal fluid drainage and continuous cerebrospinal fluid drainage by the VP shunting. A diagnosis of PRES was made due to the relative increase in cerebral perfusion pressure caused by a combination of these factors.

Fig. 2

MR imaging after cranioplasty.

The right frontal lobe showed a high signal on diffusion-weighted MR imaging and fluid attenuated inversion recovery imaging, and a high signal on the apparent diffusion coefficient map. This signal change that appeared after cranioplasty remained on postoperative days 13 and 61.

MR: magnetic resonance

Fig. 3

MR imaging after cranioplasty.

The left basal ganglia, left thalamus, and right occipital lobe showed a high signal on diffusion-weighted MR imaging and fluid-attenuated inversion recovery imaging. The right occipital lobe showed a high signal, the left basal ganglia and left thalamus showed a low signal on the apparent diffusion coefficient map. Signal changes in the right occipital lobe that appeared after cranioplasty persisted on postoperative days 13 and 61. In contrast, most signal changes in the left basal ganglia and left thalamus improved over time.

MR: magnetic resonance

Fig. 4

MR imaging after cranioplasty.

A: MR angiography revealed no new stenosis or occlusion of anterior, middle, or posterior cerebral arteries. B, C: MR venography revealed no defect in cerebral venous sinuses.

MR: magnetic resonance

When propofol was discontinued on postoperative day 2, tonic seizures appeared in the right lower extremity. Owing to circulatory depression caused by propofol, the sedative was changed to midazolam. Signal changes in the right frontal lobe and right occipital lobe that appeared after cranioplasty persisted on postoperative days 13 and 61. In contrast, most signal changes in the left basal ganglia and left thalamus improved over time (Figs. 2 and 3). The patient was extubated on postoperative day 21 because seizures did not occur after discontinuation of midazolam for 3 days. On postoperative day 169, the patient was transferred to another hospital for conservative management, with an mRS score of 5. The patient remained with right hemiparesis insufficiency, and the Glasgow Coma Scale was E1V3M6.

Discussion

We encountered a case of PRES arising after cranioplasty and VP shunting in the chronic phase of decompressive craniectomy for extensive cerebral infarction. The case was inconsistent with classic PRES in the following 3 points: 1) there was no sudden increase in blood pressure, 2) the lesion was not confined to the occipital lobe, and 3) the course was irreversible. Nevertheless, a diagnosis of PRES was made based on available clinical findings.

Recent reports have described atypical cases of PRES, including patients without episodes of rapid blood pressure elevation, those in which the lesions are not confined to the occipital lobe, and those that are irreversible. Spinal fluid drainage due to VP shunting or spinal anesthesia may cause a sudden decrease in intracranial pressure, which could cause a relative increase in cerebral perfusion pressure, resulting in the development of PRES.⁴^-⁸⁾ Thus, a rapid increase in blood pressure is not a necessary factor for the development of PRES. Approximately 75% of the patients diagnosed with PRES had lesions in the frontal and temporal lobes,⁹^,¹⁰⁾ and 10%-20% have permanent neurological sequela.¹⁰^-¹⁵⁾ PRES is characterized by vasogenic edema but may have mixed cytotoxic edema reflecting cerebral ischemia.¹¹⁾ It may follow an irreversible course due to intracranial hemorrhage, extensive brain edema, and cerebral herniation. In the current case, the signal changes in the right frontal lobe and right occipital lobe indicated vasogenic edema, while those in the left basal ganglia and left thalamus indicated cytotoxic edema. Most of them improved, but some persisted. The current case, consistent with these previously reported cases of PRES, was diagnosed accordingly.

Nevertheless, the pathogenesis and treatment of PRES have not yet been established. No randomized controlled trial on treating this syndrome has been conducted, partly due to the absence of diagnostic criteria. While cases with various causes and pathological conditions have been reported, they all have one thing in common: endothelial damage to cerebral blood vessels due to a certain cause leads to vasogenic edema.²⁾ The following factors are considered essential for the diagnosis of PRES in previously published cases; 1) episodes of sudden increase in cerebral perfusion pressure (increase in blood pressure or decrease in intracranial pressure) or the presence of cytokine-induced vascular endothelial damage due to autoimmune diseases, sepsis, renal failure, etc., 2) the presence of vasogenic edema on MR imaging, and 3) the presence of unexplained diseases other than PRES, such as cerebral infarction or cerebral arteriovenous malformation.

Even though 11 months had passed since the decompressive craniectomy, the patient's brain remained distended, which may have been associated with the development of PRES. The authors have encountered numerous cases in which VP shunting was performed simultaneously with cranioplasty, but none of these patients developed PRES. Most patients who undergo decompressive craniectomy often develop a depressed skull defect in the chronic phase, which can occasionally lead to sinking skin flap syndrome.¹⁶⁾ However, in the present case, the territory of the infarcted right middle cerebral artery appeared cystic, reflecting brain tissue necrosis and malacia. This region contained abundant fluid and remained distended (Fig. 1A-C). Although ventricular enlargement was mild (Evans index was 25%), VP shunting was performed concurrently with cranioplasty to reduce brain swelling. In such cases, if the brain remains distended during the chronic phase, draining cerebrospinal fluid during cranioplasty may increase the risk of developing PRES.

The development of PRES may not have been solely attributed to cerebrospinal fluid drainage during VP shunting; cranioplasty itself may have also contributed to its onset. Previous reports have described cases of PRES due to cerebrospinal fluid hypovolemia during shunting.¹⁷⁾ In this case, the large volume of cerebrospinal fluid drained during surgery could have been a contributing factor. Additionally, studies using CT perfusion imaging have shown changes in cerebral blood flow before and after cranioplasty in all patients, with alterations observed not only on the affected side but also on the contralateral side.¹⁸⁾ In this case, CT perfusion imaging was not performed; however, significant changes in overall cerebral blood flow may have occurred as a result of the surgery. Although PRES has not been reported to occur solely due to cranioplasty, a combination of factors, such as changes in cerebral blood flow and a reduction in intracranial pressure due to cerebrospinal fluid drainage, may overwhelm the brain's autoregulatory capacity and contribute to the development of PRES.

This report has several limitations. The first one is the accuracy of the diagnosis of PRES. Differential diagnoses include cerebrovascular disorders, such as acute cerebral infarction and cerebral venous sinus thrombosis. Acute cerebral infarction was ruled out in this case because there was no stenosis or occlusion of the cerebral vessels on MR angiography. Even if stenosis or occlusion had recanalized, the lesion extended into multiple cerebral vascular territories, and the source of the embolus could not be identified using electrocardiography or ultrasonography. Cerebral venous sinus thrombosis was also ruled out in this case, because the venous sinus was normally visualized in MR venography. Pseudohypoxic cerebral swelling was caused by rapid loss of cerebrospinal fluid, similar to this case, but was also ruled out because there was no venous congestion or dilation on MR venography.¹⁹⁾ Based on these points, we considered the diagnosis of PRES as the most appropriate. Second, how to prevent this condition remains unclear. As mentioned, a two-staged approach involving VP shunting and cranioplasty may be an option for skull defects complicated by brain distension. However, the success of separate procedures is not guaranteed since PRES can develop even when a VP shunt is performed alone. Thirdly, it cannot be definitively concluded that cranioplasty is directly associated with the onset of PRES. The main cause of the development of PRES may have been the drainage of cerebrospinal fluid during surgery, rather than being directly caused by cranioplasty. If brain blood flow changes before and after surgery had been assessed using modalities such as CT perfusion imaging, single-photon emission computed tomography, or arterial spin labeling MR perfusion imaging, it might have been possible to establish a causal relationship between cranioplasty and PRES.

Conclusion

When cranioplasty is performed simultaneously with VP shunting for a skull defect complicated by brain distension, the sudden drop in intracranial pressure may lead to the development of PRES.

Informed Consent

We obtained informed consent for publication from the patient.

Conflicts of Interest Disclosure

All authors have no conflict of interest.

References

責任著者(Corresponding author)

J-STAGEへの登録はこちら（無料）