Peri-ictal Hyperperfusion Revealed by Arterial Spin Labeling Perfusion Images in 2 Patients with Post-aneurysmal Subarachnoid Hemorrhage Epilepsy and Associated Old Infarction in the Middle Cerebral Artery Territory

Takato MORIOKA; Satoshi INOHA; Akifumi YOKOMIZO; Takafumi SHIMOGAWA; Nobutaka MUKAE; Kazuhisa KUWABARA; Ayumi YONAMOTO; Masayuki YANAGIDA; Fumihito MUGITA; Ayumi SAKATA; Hiroshi SHIGETO; Koji YOSHIMOTO

doi:10.2176/jns-nmc.2026-0116

Abstract

To make a pathophysiological diagnosis of post-aneurysmal subarachnoid hemorrhage epilepsy, which has complex pathologies, in a neuro-emergency setting, peri-ictal magnetic resonance imaging including arterial spin labeling perfusion images was conducted in 2 patients with associated old infarction due to vasospasm in the middle cerebral artery territory. In patient 1, who had a localized infarction and no apparent circulatory disturbance in the middle cerebral artery territory on interictal magnetic resonance imaging, focal to bilateral tonic–clonic seizures developed twice within 1 year after the subarachnoid hemorrhage onset. In both events, peri-ictal hyperperfusion was observed in a tight anatomical relationship with the infarction. In patient 2, who had extensive infarction and circulatory disorders in the middle cerebral artery area, the fourth focal to bilateral tonic–clonic seizures developed 27 years after the subarachnoid hemorrhage. In this event, peri-ictal hyperperfusion was observed predominantly in the ipsilateral frontal interhemispheric cortex in the anterior cerebral artery territory, slightly away from the infarction. Further case observations with peri-ictal magnetic resonance imaging, in addition to electroencephalography, would help to clarify the complex pathologies of post-subarachnoid hemorrhage epilepsy.

Introduction

Epilepsy is a well-recognized sequela of subarachnoid hemorrhage (SAH) after ruptured aneurysm, with an incidence ranging from 4.2% to 26%.¹⁾ Post-SAH epilepsy is more common in young patients with high SAH grades and middle cerebral artery (MCA) aneurysm.¹^,²⁾ Furthermore, coexisting parenchymal pathologies that can become epileptogenic lesions in the future, including intracerebral hemorrhage (ICH) at the time of SAH onset and subsequent cerebral infarction due to vasospasm or surgery, all complicate the pathogenesis of post-SAH epilepsy.¹^,²⁾ Probably owing to the complex pathologies, Zhou et al.³⁾ recently reported that as a cause of post-stroke epilepsy, SAH is more common than hemorrhagic and ischemic stroke. However, the exact pathophysiological mechanism of post-SAH epilepsy has not been indicated. Given post-SAH epilepsy can worsen clinical outcomes and negatively affect quality of life,¹⁾ as with other post-stroke epilepsy, it is important to promptly and accurately diagnose the pathophysiology at a neuro-emergency setting.⁴⁾

However, at most acute care hospitals in Japan, including our facility, routine electroencephalography (EEG), which is traditionally used to diagnose epilepsy, is only available during consultation hours on weekdays. This commonly causes delays in the timing of diagnosis.⁴^-¹²⁾ Furthermore, the lack of 24-hour EEG specialists onsite at all acute care hospitals is likely to hinder prompt diagnosis using EEG.¹⁰⁾ To overcome these limitations, we added arterial spin labeling (ASL) perfusion imaging to routine magnetic resonance imaging (MRI) because this technique is now available 24 hours a day at many acute care hospitals owing to the development and widespread use of acute stroke treatment.⁴^-¹²⁾ We previously reported that the capture of ictal or peri-ictal ASL hyperperfusion, which arises secondary to increased metabolism owing to excessive neuronal excitation, reveals a tight anatomical relationship with epileptogenic lesions, making it useful for the prompt and correct diagnosis of structural focal epilepsy.⁴^-¹²⁾ Similar results were reported from other facilities.¹³^-¹⁸⁾ When diagnosing post-stroke epilepsy, it is important to compare it with interictal MRI, including magnetic resonance angiography (MRA), to evaluate the underlying circulatory disturbances.⁴⁾

For 2 years, from January 2024 to December 2025, 88 seizures were captured in 72 patients using both 1.5-Tesla pseudo-continuous ASL with triple post-labeling delays (PLDs) of 1.5, 1.75, and 2.0 seconds (s), and EEG as was previously described.⁴^,⁸^-¹²^,¹⁹^-²²⁾ We selected 2 patients with post-SAH epilepsy. Both patients had ruptured MCA aneurysm and associated old infarction due to subsequent vasospasm in the ipsilateral MCA territory. Herein, we report the findings of MRI and EEG performed during the interictal and peri-ictal states of these 2 patients and discuss the role of their infarctions in the development of post-SAH epilepsy.

Case Reports

Patient 1

In the male patient, SAH developed owing to a ruptured right MCA aneurysm and subsequent localized infarction due to a vasospasm in the right frontal lobe at the age of 46 years (Fig. 1A, yellow arrows). Four months after clipping surgery, the patient returned home with a modified Rankin Scale (mRS) score of 1. However, he experienced his first and second focal to bilateral tonic–clonic seizures (FBTCSs) 10 months and 1 year after SAH onset, respectively, and was immediately transported to our hospital.

Figure 1

(Patient 1).

(A-D) FLAIR (A), pseudo-colored original ASL (B) and ASL fused with FLAIR at PLDs of 1.5, 1.75, and 2.0 s (C), and MRA (D) during the interictal state.

(E-J) Original ASL (E) and ASL fused with FLAIR at PLDs of 1.5, 1.75, and 2.0 s (F), DWI (G), subtracted ASL images of interictal from peri-ictal ASL images (H) and SIACOM with FLAIR at PLDs of 1.5, 1.75, and 2.0 s (I), and MRA (J) performed 50 min after the admission for the first seizure.

(K) EEG with Av reference performed 4.5 hrs after the first admission.

(L-Q) Original ASL (L), ASL fused with FLAIR (M), subtracted ASL (N), and SIACOM with FLAIR at PLDs of 1.5, 1.75, and 2.0 s (O), DWI (P), and MRA (Q) performed 1 hr after admission for the second seizure. White arrows and white dotted arrows indicate the hyperperfusion area and the surrounding hypoperfusion area, respectively.

(R) EEG with Av reference performed 32 hrs after the second admission.

The text presents details.

ASL: arterial spin labeling; Av: averaged; DWI: diffusion-weighted image; EEG: electroencephalography; FLAIR: fluid-attenuated inversion recovery; hr (s): hour (s); MRA: magnetic resonance angiography; min: minutes; PLD: post-labeling delay; s: second; SIACOM: subtraction of (peri-)ictal-interictal ASL images co-registered with conventional magnetic resonance images

Although interictal MRI was conducted 11 days after the first seizure, these MRI findings are described first, to understand the interictal hemodynamic state of this patient. ASL was imaged as previously described.⁴^,⁸^-¹²^,¹⁹^-²²⁾ The pseudo-colored original ASL (Fig. 1B) and ASL images fused with fluid-attenuated inversion recovery (FLAIR) (Fig. 1C) revealed defective ASL signals in all PLDs (Fig. 1B and C; white arrows). However, the ASL signal in the MCA territory was not reduced (Fig. 1B and C). MRA showed no differences in the visualization of the bilateral MCAs (Fig. 1D). T2*-weighted imaging (T2*WI) failed to reveal any superficial cortical siderosis.

MRI at the time of the first seizure was conducted 50 minutes after the patient was transported outside regular consultation hours. On arrival, the patient continued to experience FBTCS, requiring the intravenous administration of 7.5 mg diazepam (DZP). MRI was performed when his consciousness level was II-20 on the Japan Coma Scale (JCS). Original ASL (Fig. 1E) and ASL fused with FLAIR (Fig. 1F) at a PLD of 1.5 s clearly showed prominent hyperperfusion around the infarct in the right frontotemporal lobes, particularly in the posterior part of the infarct (Fig. 1E and F, white arrows). Furthermore, the signal intensity around the area of hyperperfusion was lower than that on the contralateral side (Fig. 1F, white dotted arrows). This increased signal intensity slightly decreased in intensity and extent at a PLD of 1.75 s, but remained intense even at a PLD of 2.0 s. A faint high signal intensity was observed in the corresponding area on diffusion-weighted imaging (DWI) (Fig. 1G, yellow arrows). Subtracted images (Fig. 1H) and subtraction of the (peri-) ictal from interictal ASL images co-registered with anatomical MRI (SIACOM), which was created as previously described,⁴^,⁸^,⁹^,¹¹^,¹⁹^,²¹^,²³⁾ largely eliminated any background ASL signals from the original ASL, thereby more clearly revealing the tight topographical relationship between the infarct and the area of hyperperfusion (Fig. 1I). MRA revealed no left-right difference in the visualization of the peripheral MCAs, as in the interictal period (Fig. 1J). Intravenous 500 mg levetiracetam (LEV) was subsequently administered.

In the state of II-20 on JCS, EEG conducted 4.5 hours after transport revealed periodic discharges (PDs) of approximately 1.7 Hz in the right temporo-frontal regions, with the maximal amplitude observed at the right posterior temporal region (T6) (Fig. 1K). The subsequent day, the patient regained full consciousness, and oral administration of 1,000 mg LEV was initiated.

A second FBTCS developed in the patient during the transition to lacosamide (LCM) therapy. He was transported over the weekend, and 5 mg DZP was intravenously administered. MRI was conducted 1 hour after transport in state II-20 on the JCS. Original ASL (Fig. 1L), ASL fused with FLAIR (Fig. 1M), subtracted ASL (Fig. 1N), and SIACOM (Fig. 1O) all showed cerebral blood flow (CBF) distribution abnormalities, similar to those observed in the peri-ictal ASL at the time of the first seizure. However, ASL after the second seizure showed fewer background signals and more pronounced peri-infarct hyperperfusion than ASL after the first seizure. Furthermore, as observed at the time of the first seizure, DWI exhibited a faint high signal intensity in areas corresponding to those where hyperperfusion remained visible on ASL at a PLD of 2.0 s (Fig. 1P, yellow arrows). MRA additionally showed no significant left-right difference in the visualization of the MCA periphery (Fig. 1Q). After the MRI examination, 100 mg LCM was administered intravenously. The following day, the patient regained consciousness, and the dose of oral LCM was increased from 50 to 100 mg.

EEG performed 32 hours later revealed high-amplitude slow waves in the right temporal region, with a maximal amplitude at T6 (Fig. 1R). The dose of LCM was gradually increased to 300 mg. The patient has thus far remained seizure-free for 8 months.

Patient 2

A 64-year-old woman underwent bilateral fronto-temporo-parietal craniotomies and clipping of ruptured right MCA and unruptured left MCA aneurysms at the age of 37 years. MRI revealed extensive infarction in the right frontoparietotemporal lobes due to vasospasm and the left frontal lobe due to surgery (Fig. 2A and B). Neurological examination revealed left hemiparesis and dysarthria, with an mRS score of 3. Nine months after surgery, the patient began experiencing FBTCS, which was controlled once every few years with 1,000 mg valproic acid or 1,000 mg LEV. However, she was subsequently admitted to our hospital with a fourth FBTCS, beginning in the left upper limb, which developed 27 years after SAH onset.

Figure 2

(Patient 2).

(A-E) FLAIR (A), DWI (B), original ASL (C) and ASL fused with FLAIR at PLDs of 1.5, 1.75, and 2.0 s (D), and MRA (E) during the interictal state.

(F) EEG with Av reference performed 1.5 hrs after the admission.

(G-L) Original ASL (G), ASL fused with FLAIR (H), subtracted ASL images of interictal from peri-ictal ASL images (I) and SIACOM with FLAIR at PLDs of 1.5, 1.75, and 2.0 s (J), DWI (K), and MRA (L) performed 6 hrs after admission.

The text presents details.

ASL: arterial spin labeling; Av: averaged; DWI: diffusion-weighted image; EEG: electroencephalography; FLAIR: fluid-attenuated inversion recovery; hrs: hours; MRA: magnetic resonance angiography; PLD: post-labeling delay; s: second; SIACOM: subtraction of (peri-)ictal-interictal ASL images co-registered with conventional magnetic resonance imaging

Although interictal MRI was conducted 2 months after the seizure episode, these MRI findings are described first. In both original ASL (Fig. 2C) and the ASL images fused with FLAIR (Fig. 2D), the right infarction appeared signal-free, except for the residual tissue at all PLDs (Fig. 2C and D; white arrows). Signal intensity in the right hemisphere, particularly in the region surrounding the right MCA, was markedly reduced. On MRA, visualization of the frontal branch of the right MCA was poorer than that of the left MCA (Fig. 2E, yellow arrows). Furthermore, clip artifacts were observed in both MCAs. T2*WI did not reveal any superficial cortical siderosis.

Immediately after transport, within the consultation hour, the patient presented with FBTCS accompanied by left conjugate deviation, and 5 mg DZP was administered intravenously. In the state of II-20 on JCS, EEG performed 1.5 hours after the transport revealed PDs of approximately 1 Hz in the right frontal region, centered on the right frontal pole (Fig. 2F, red arrows), which increased in frequency and ultimately evolved into ictal discharges (Fig. 2F, blue arrows). These ictal discharges spread to the left frontal pole, developing a spike-and-wave morphology (Fig. 2F, blue line) and having further evolution, including an increase in frequency in the right frontal region (Fig. 2F, red lines). The patient was diagnosed with a transition from FBTCS to non-convulsive status epilepticus, and 5 mg DZP was administered intravenously again, followed by intravenous administration of 100 mg LCM.

In the state of II-10 on JCS, peri-ictal MRI was conducted 6 hours after transport. At PLD of 1.5 s, the original ASL (Fig. 2G) and ASL images fused with FLAIR (Fig. 2H) showed hyperperfusion in both frontal interhemispheric cortices and the left lateral frontal cortex in the anterior cerebral artery (ACAs) territories. At 1.75 s, hyperperfusion decreased in magnitude and extent. At 2.0 s, hyperperfusion persisted in the right frontal interhemispheric cortex but only minimally in the left frontal lobe. On the subtracted image (Fig. 2I) and SIACOM (Fig. 2J), hyperperfusion was prominent in the right interhemispheric cortex and very mild in the left frontal lobe. On DWI, a faint increase in the signal was suspected in the right frontal interhemispheric cortex, where ASL hyperperfusion persisted for 2.0 s (Fig. 2K, yellow arrow). MRA showed poor visualization of the frontal branches of the right MCA compared with those of the left MCA, similar to the interictal findings (Fig. 2L).

The patient regained full consciousness the next day, and additional oral administration of LCM was initiated. She has thus far remained free of seizures during these 2 years with 1,000 mg LEV and 200 mg LCM.

Discussion

In patient 1, who had a localized infarction and no apparent circulatory disturbance in the MCA territory on interictal MRI, prominent ASL hyperperfusion was consistently observed in a tight anatomical relationship with the infarction after both FBTCS events, also indicating the high consistency of the ASL examination itself.²⁴⁾ The surrounding area showed reduced signals owing to blood stealing, as previously reported.⁴^,⁷^,¹⁰^,¹¹⁾ Furthermore, the area that remained highly hyperperfused at a PLD of 2.0 s was the most epileptically activated and showed an increased metabolism.⁴^-⁶^,⁹^,¹¹⁾ It has been explained that when intense seizures persist for a long duration, in addition to excessive release of excitatory amino acids and an influx of calcium ion into cells, CBF is insufficient to compensate for the increased metabolism; in other words, uncoupling between CBF and metabolism (relative ischemia) occurs, generating cytotoxic edema, which causes the area to show a high signal on DWI.⁴^-⁶^,⁹^,¹¹⁾ A subsequent EEG revealed paroxysmal discharges that corresponded closely to the hyperperfused area. However, as mentioned in our previous reports,⁴^-¹²^,²³⁾ ASL, particularly SIACOM, more clearly depicts the pathophysiology of structural focal epilepsy, in which the infarction is an epileptogenic lesion.

In contrast, in patient 2, who had extensive infarction and circulatory disorders in the MCA territory, no peri-ictal ASL hyperperfusion was detected near the infarction. Although the original ASL revealed a peri-ictal hyperperfusion in the bilateral frontal lobe, SIACOM clearly indicated it localized mainly at the ipsilateral frontal interhemispheric cortex in the ACA territory, without any apparent epileptogenic lesion, slightly away from the infarction. Furthermore, this area was suspected to have cytotoxic edema, on the basis of DWI. This cytotoxic edema, similarly to that in patient 1, is believed to have resulted from intense and prolonged seizure activities in this area.⁴^-⁶^,⁹^,¹¹⁾ This distribution of peri-ictal hyperperfusion with cytotoxic edema explains that the ictal EEG discharges at the contralateral frontal pole were recorded through the mechanism of "paradoxical lateralization"; that is, the activities located in the interhemispheric cortex project transversely or parallel (not perpendicular) to the contralateral scalp surface.²⁵⁾ This result further indicates the usefulness of assessing both the CBF and electrophysiology perspectives to understand the pathophysiological conditions of epilepsies.⁴^-¹²^,²³^,²⁴⁾

The mechanism underlying ictal EEG discharge and peri-ictal ASL hyperperfusion in the ACAs territory, slightly distant from the infarct in the right MCA territory, is unknown. We reported that in patients with post-ICH epilepsy who had circulatory disturbances in the MCA territories, probably due to regression of the MCA occurring long after ICH onset, in addition to the primary damage to the MCA at onset, ictal ASL hyperperfusion may not be maintained and visualized during the peri-ictal state.⁴^,²⁶⁾ Patient 2 had experienced this condition 27 years before the current seizure episode. We speculate that the epileptic activities at this time, initially occurring near the infarct, spread to the frontal interhemispheric cortex with secondary generalization and remained intense, receiving sufficient blood supply from the ACA during the peri-ictal period, whereas the initial ictal activities and associated hyperperfusion in the MCA territory could not be maintained during the peri-ictal period owing to insufficient blood flow from the MCA.

In conclusion, further experience with peri-ictal MRI, in addition to EEG, would help to clarify the complex pathologies of post-SAH epilepsy.

Disclaimer

Author Koji Yoshimoto is one of the Editorial Board members of the Journal. This author was not involved in the paper-review or decision-making process for this paper.

Acknowledgments

We thank Ryoji Shiraki and colleagues at the Hachisuga Hospital for supporting this study. We thank Editage for editing the manuscript.

Author Contributions

Takato Morioka, Satoshi Inoha, and Akifumi Yokomizo: Study design and project outline. Takato Morioka, Kazuhisa Kuwabara, Ayumi Yonamoto, Masayuki Yanagida, Fumihito Mugita, and Ayumi Sakata: Data acquisition, analysis, and manuscript drafting. Takafumi Shimogawa, Nobutaka Mukae, Hiroshi Shigeto, and Koji Yoshimoto: Critical revision of the manuscript.

Conflicts of Interest Disclosure

All authors have no conflict of interest.

Statement of Ethics

The authors confirm that written informed consent for publication of this report was obtained from both patients. Ethical approval was obtained from the institutional review board of the Hachisuga Hospital (No. 22-1).

References

Corresponding author

Register with J-STAGE for free!