Post-traumatic Transient Neurological Dysfunction: A Proposal for Pathophysiology

Abstract

Unexplained neurological deterioration is occasionally observed in patients with traumatic brain injuries (TBIs). We aimed to describe the clinical features of post-traumatic transient neurological dysfunction and provide new insight into its pathophysiology. We retrospectively collected data from patients with focal neurological deterioration of unknown origin during hospitalization for acute TBI for 48 consecutive months. Brain imaging, including computed tomography, diffusion-weighted imaging and perfusion-weighted imaging, and electroencephalography were conducted during the episodes. Fourteen (2.0%) patients experienced unexplained focal neurological deterioration among 713 patients who were admitted for traumatic intracranial hemorrhage during the study period. Aphasia was the predominant symptom in all patients, and hemiparesis or hemianopia was accompanied in three patients. These symptoms developed within 14 days after trauma. Structural imaging did not show any significant interval change, and electroencephalography showed persistent arrhythmic slowing in the corresponding hemisphere in most patients. Perfusion imaging revealed increased cerebral blood flow in the symptomatic hemisphere. Surgical intervention and anti-seizure medications were ineffective in abolishing the symptoms. The symptoms disappeared spontaneously after 4 h to 1 month. Transient neurological dysfunction (TND) can occur during the acute phase of TBI. Although TND may last longer than a typical transient ischemic attack or seizure, it eventually resolves regardless of treatment. Based on our observation, we postulate that this is a manifestation of spreading depolarization occurring in the injured brain, which is analogous to migraine aura.

Introduction

Neurological deterioration occurs in 3–24% of mild head injuries that are complicated by intracranial hemorrhage.¹ Post-traumatic neurological deterioration is mostly attributed to an expansion of intracranial lesions² and potential vasospasm,³ but deterioration without any relevant structural change is occasionally encountered in clinical practice. However, its nature is not understood and is rarely discussed.

There have been a few reports describing transient neurological dysfunction (TND) of unknown cause after concussion. Most of those studies were published before 2000 and were based on brain computed tomography (CT), which is insufficient to disclose its pathophysiology.^4,5 In traumatic brain injury (TBI) with intracranial hemorrhage, TND has been under-recognized because neurological deterioration is often believed to be caused by brain swelling and compression associated with the traumatic lesion, and the resolution of symptoms could be misinterpreted as being due to the spontaneous regression of TBI or surgery.⁶

We experienced the episodes of unexplained TND in patients with acute TBI with intracranial hemorrhage. After a thorough evaluation to rule out potential causes and to elucidate its pathophysiology, examination of therapeutic trials, and observation of the clinical courses in a series of patients, we could delineate the unique features of post-traumatic TND.

We hypothesize that spreading depolarization (SD) in response to noxious stimuli in the cerebral cortex is responsible for post-traumatic TND, as the cortical spreading depression elicited by focal stimulation of the rabbit cerebral cortex in Leão’s experiment has been suggested as a pathophysiology of migraine aura.^7,8

In this study, we describe the frequency, characteristics, course, and outcome of post-traumatic TND and infer its pathophysiology.

Methods

We collected data from patients who had focal neurological deterioration of unknown origin or discordant to the lesion during admission for acute TBI in the Neurosurgery Department of Kangwon National University Hospital, a regional teaching hospital, between February 2014 and January 2018.

Patients with intracranial hemorrhage were initially admitted to the intensive care unit (ICU) for at least 48 hours if the ICU was available. Neurological examination, including the Glasgow Coma Scale, pupil reactivity, and motor power, was performed every hour in the ICU, and health status including conscious level and vital sign was checked every 4 h in the general wards. Prophylactic anti-seizure medication (ASM) was administered since presentation in all patients with intracranial hemorrhage. Levetiracetam was the first-line agent, and valproic acid was used in those with renal insufficiency. Large hematomas associated with neurological deterioration or raised intracranial pressure (midline shift ≥5 mm or compression of the basal cistern) were surgically removed.

If the attending neurosurgeons determined that the patients had unexplained neurological deterioration, they consulted a study neurologist (S.-Y.L.), who is experienced in neurophysiology and critical care, to further assess possible causes, including ischemia, seizure, medication, and systemic derangement.

Brain CT was performed for all patients, and diffusion-weighted magnetic resonance (MR) imaging (DWI) was also included as the initial evaluation for unexplained neurological deterioration from August 2015. Beginning in August 2016, MR angiography (MRA) using a three-dimensional time-of-flight technique and perfusion-weighted MR imaging (PWI) using the dynamic susceptibility contrast technique were performed if patients’ conditions allowed the examination. Scalp electroencephalographies were recorded over all brain areas using 21 electrodes according to the conventional 10–20 system of electrode placement for 15–30 min.

Spectral analysis of electroencephalography was performed to display the topographic abnormality associated with TND. The data preprocessing and analysis were performed in MATLAB r2002a (MathWorks, Inc., Natick, MA). Continuous wavelet transform (CWT) decomposes the signal, $x (t)$ , into time-frequency space, $W (τ, s)$ by convolution of the wavelet function. $W (τ, s) = \int_{- \infty}^{\infty} x (t) ψ_{s} (t - τ) d t = f (t) * ψ_{s} (t)$

The Morlet wavelet resembles the oscillation kernel modulated with a Gaussian envelope and is commonly adapted as a mother wavelet function. By contracting and dilating the mother wavelet with different scales, the scaled wavelet can represent different frequency oscillation and window sizes corresponding to scales (i.e., $ω_{0}$ represents nature frequency and $s$ indicates the wavelet scale). $ψ_{s} (t) = ψ (\frac{t}{s}) = \frac{1}{\sqrt{s}} π^{\frac{- 1}{4}} e^{- \frac{1}{2} {(\frac{t}{s})}^{2}} e^{i ω_{0} \frac{t}{s}} \begin{matrix} F \\ \leftrightarrow \end{matrix} \hat{ψ_{s} (ω)} = \sqrt{s} \hat{ψ (s ω)} = \sqrt{s} π^{\frac{- 1}{4}} e^{- \frac{1}{2} {(ω_{0} - s ω)}^{2}}$

The center frequency of each wavelet function $f_{s - c}$ has an intuitive relationship to scale. $F$ operates the Fourier transform between time and frequency domain. $f_{s - c} = \frac{ω_{0}}{s}$

The CWT spectrum of electroencephalography could be calculated as follows: $W_{EEG} (t, s) = EEG (t) * ψ_{s} (t)$

To calculate the power spectrum density of a specific frequency band, the power of each frequency bins was summed up and averaged over time respectively. We calculated the median frequency of each channel and constructed its topology mapping. The average frequency spectrum of each hemisphere was derived by averaging the power of the same corresponding frequency bins in overall channels within the hemisphere.

To identify the consecutive patients who had unexplained neurological deterioration for this study, the consultation records of the study neurologist were screened. After reviewing whole medical records of TBI hospitalization, the patients were included only when they had focal neurological deterioration and other potential diagnoses were ruled out. We recorded the details of the index episodes, including the symptoms, course, outcome, and therapeutic trials and their effects. Clinical characteristics, including the neurological status at presentation and before index events, the type of TBI, the treatment of TBI, and the occurrence of seizures, were also collected from medical records.

The number of overall patients admitted for TBI was assessed to estimate the incidence of post-traumatic TND. Admission for TBI was identified by diagnostic code S06 (intracranial injury) according to ICD-10 as the primary diagnosis at discharge.

This study was approved by the institutional review board of Kangwon National University Hospital (KNUH-2021-04-020). The requirement of informed consent was waived owing to the retrospective design.

Results

During the study period, there were 839 admissions for TBI, including intracranial hemorrhage in 713 (extradural hemorrhage [EDH] in 81, subdural hemorrhage [SDH] in 421, subarachnoid hemorrhage [SAH] in 107, and contusion in 120), skull fracture in 42, and concussion in 84 patients. They were 535 males and 304 females with ages ranging from 25 days to 97 years (median 64 years; interquartile range [IQR] 47–76 years). Among them, 17 adults had significant neurological deterioration that we could not identify attributable causes. Among these patients, three who had nonfocal symptoms, such as behavioral changes and hypersomnolence, were excluded because there is a possibility that delirium or encephalopathy associated with systemic infection was not completely ruled out. Overall, 14 adult patients (1.7% of all TBI, 2.0% of traumatic intracranial hemorrhage, and 3.3% of SDH) were included in this study. They were 7 males and 7 females with ages ranging from 40 to 83 years.

We first present three representative cases and then summarize the clinical data of 14 patients with post-traumatic TND.

Illustrative case 1 (patient 2)

A 77-year-old woman presented with sudden headache and vomiting that lasted for 3 days. She was alert without neurological deficits. Brain CT showed acute SDH in the left convexity with 10 mm thickness and a 10 mm midline shift. She underwent craniotomy, and a subdural hematoma was removed. Ten days later, she developed word-finding difficulty. Brain CT showed a small amount of hemorrhage and fluid collection (8 mm) in the left subdural space with swelling of the underlying cortex, which was unchanged from the previous imaging conducted 4 days ago (Fig. 1A). She became mute over 2 days. Electroencephalography performed on the third day after the onset of symptoms revealed continuous low amplitude arrhythmic theta to delta slowing over the left hemisphere without epileptiform discharge (Fig. 1B). She improved gradually from the fourth day of speech disturbance. Overall, her speech disturbance lasted for 6 days. Brain MR imaging and MRA conducted after improvement showed mild stenosis in the left proximal internal carotid artery and both middle cerebral arteries without brain parenchymal lesions.

FIG. 1.

Brain CT and electroencephalography from patient 2. Brain CT performed 10 days after presentation when motor aphasia developed showed small amount of hemorrhage and fluid collection in the left subdural space with the obliteration of underlying sulci, which was unchanged from the previous imaging conducted 4 days before (A). Electroencephalography performed on the third day of the symptom revealed continuous attenuation with arrhythmic theta to delta slowing over the left hemisphere and no epileptiform discharge (B). CT, computed tomography.

Illustrative case 2 (patient 9)

A 47-year-old woman had headache that lasted for 3 days after falling down the stairs and developing a head injury. Brain CT showed hemorrhagic contusion in both the orbitofrontal (0.6 cc) and the left anterior temporal lobe (8 cc) with an SDH along the left cerebral convexity (2 mm) and falx (1 mm) with a 4 mm midline shift. Surgical intervention was not performed, and mannitol and levetiracetam were administered. Ten days after TBI, she developed a speech disturbance, and brain CT showed worsening of the left hemispheric swelling with progression of midline shift (7 mm). The patient underwent frontotemporoparietal craniotomy to evacuate the acute SDH and temporal contusional hemorrhage. Postoperative brain CT revealed sufficient removal of the hematoma, with improved midline shifting. However, her speech gradually worsened, and global aphasia developed on the next day. She was alert and cooperative, and no other neurological signs were observed. Brain CT (Fig. 2A) and DWI showed a decrease in the pressure effect and no significant change that could explain the symptom deterioration. Electroencephalography showed persistent high-amplitude arrhythmic delta activity over the entire left hemisphere and no epileptiform discharge (Fig. 2B). Intravenous valproic acid was administered but was not helpful. She improved gradually over approximately 3 weeks from the fourth day after the onset of speech disturbance. She had no sequelae, and no additional lesion was observed on brain CT. She had a small amount of theta slowing in the left temporal area on electroencephalography at the 1-year follow-up.

FIG. 2.

Brain CT and electroencephalography conducted when global aphasia developed on the day after craniotomy (10 days after trauma) from patient 9. Brain CT showed contusion in both orbitofrontal and the left anterior temporal lobe with relieved pressure effect after surgery (A). Electroencephalography showed persistent high arrhythmic delta activity over the entire left hemisphere and no epileptiform discharge (B). CT, computed tomography.

Illustrative case 3 (patient 11)

An 83-year-old woman had headache and back pain after she slipped and fell. SDH along the falx (15 mm) and left parietotemporal area (4 mm) and multiple spinal compression fractures at T12 and L1 were observed. She was treated conservatively and received prophylactic levetiracetam. Thirteen days after TBI, she had difficulty with verbal expression and lifting her right arm. The next day, global aphasia, monoplegia of the right arm, and hypersomnolence developed. Brain CT showed obliteration of the sulci underneath the SDH with only chronologic changes in its density from the fourth day after trauma (Fig. 3A). DWI and MRA revealed no additional abnormalities that could have caused the symptoms. PWI revealed an increase in cerebral blood flow (CBF) in the left parietotemporal area (Fig. 3B). Electroencephalography showed persistent medium amplitude arrhythmic theta to delta slowing over the left hemisphere (Fig. 3C). There was no response to a therapeutic trial of oxcarbazepine followed by intravenous valproic acid. Three days after symptom onset, she spoke a few words, showing a gradual improvement. She had normal motor function but a subtle decrease in verbal fluency at 4 weeks after symptom onset. She fully recovered when she was discharged at 6 weeks after symptom onset.

FIG. 3.

Brain images and electroencephalography conducted when global aphasia and monoplegia of right arm developed 13 days after trauma from patient 11. Brain CT showed subdural hematoma along the falx and left parietotemporal area with only chronologic change in the attenuation of the hematoma from the fourth day after trauma (A). Perfusion-weighted MR imaging revealed an increase in CBF in the left parietotemporal area (B). Electroencephalography showed persistent medium amplitude arrhythmic theta to delta slowing over the left hemisphere without epileptiform discharge (C). CBF, cerebral blood flow; CT, computed tomography; MR, magnetic resonance.

Clinical phenotype

The clinical characteristics of underlying TBI and post-traumatic TND are summarized in Tables 1 and 2, respectively.

Table 1.

Baseline Characteristics and Outcomes of Patients Who Experienced Unexplained Transient Neurological Dysfunction After Traumatic Brain Injury

	Age/sex	Initial presentation; GCS	TBI lesion (thickness, mm, or volume, cc)*	Midline shift (mm)	Treatment for TBI (before TND)	Neurological status before TND; GCS	GOS at 3 months
1	52/M	Right hemiparesis (grade 4+) and dysarthria; 13	Acute EDH, SDH, and SAH in the left convexity (23)	11	Craniotomy and EDH/SDH removal	Mild dysarthria; 15	5
2	70/F	Headache and vomiting; 15	Acute SDH in the left convexity (10)	7	Craniotomy and SDH removal	Normal; 15	5
3	72/M	Right hemiparesis (grade 4) and dysarthria; 13	EDH, SDH, SAH, and skull fracture in the left convexity (3); contusion in the left lateral frontal area (1)	5	Craniotomy and EDH/SDH removal	Subtle weakness in the right arm; 15	5
4	56/M	Headache; 14	Acute (mixed stage) SDH in the left convexity (12)	8	Craniotomy and SDH removal	Normal; 15	5
5	78/F	Headache; 15	Acute SDH in the left convexity (3)	2	Conservative	Normal; 15	5
6	83/M	No neurological symptom, multiple traumas; 15	Acute EDH, SDH, and skull fracture in the left temporoparietal area (6)	0	Conservative	Normal; 15	5
7	76/M	Global aphasia and right hemianopia; 10	Acute SDH in the falx (2), small contusion in the left orbitofrontal area (0.1); small amount of SAH in the left high frontal area (1)	0	Conservative	Present with TND; 10	5^a
8	79/F	Headache; 15	Acute SDH in the falx (3) and SAH in the basal cistern and left sylvian fissure (3)	2	Conservative	Normal; 15	5
9	47/F	Headache; 15	Acute SDH in the left convexity (2) and contusion in the left temporal lobe (18)	4	Conservative	Normal; 15	5
10	60/M	Headache; 15	Acute SDH in the left convexity (16)	12	Craniotomy and SDH removal	Normal; 15	5
11	83/F	Headache; 15	Acute SDH in the falx (15) and left parietal lobe (4)	0	Conservative	Paraparesis (because of T12 and L1 compression fractures); 15	5
12	40/M	Drowsy mentality; drunken state; 11	Acute EDH and SDH in the left parieto-occipital area (12)	8	Craniotomy and EDH/SDH removal	Disorientation; normal speech; 14	5
13	47/M	Stuporous mentality; 8	Acute SDH and SAH in the left hemisphere (11), SDH in the falx (2) and contusion in left orbitofrontal area (5)	6	Craniectomy and SDH removal	Dysphasia; 13	4
14	62/F	Headache; 15	Lt. tentorial hemorrhage (8)	0	Conservative	Dysarthria, right leg weakness as sequelae of past stroke; 14	4

^*EDH, SDH, and SAH were measured by thickness (mm), and contusion was measured by volume (cc) at presentation, except for patient 12, in whom the size was measured when the lesion significantly progressed and needed surgical treatment.

GOS measured at 1 month, before acute ischemic stroke occurred.

EDH, extradural hemorrhage; GCS, Glasgow Coma Scale; GOS, Glasgow Outcome Scale; SAH, subarachnoid hemorrhage; SDH, subdural hemorrhage; TBI, traumatic brain injury; TND, transient neurological dysfunction.

Table 2.

Symptoms and Laboratory Findings of Post-traumatic TND and Treatments Conducted for TND

	Age/sex	TND symptom; GCS	Onset of TND after trauma	Time to peak/time to improving	Duration of TND	Electroencephalography	Brain CT	DWI	Vascular imaging	PWI	Treatment
1	52/M	Motor aphasia; 12	9 days	<1 day/3 days	27–40 days	Low amplitude ATDA in the LH	NC	NC	n	n	Observation
2	70/F	Motor aphasia; 11	13 days	2 days/4 days	6 days	Low amplitude ATDA in the LH	NC	n	Mild stenosis at the left proximal ICA and both MCA	n	Observation
3	72/M	Global aphasia/right hemiparesis (grade 3); 11	2 days	1 day/2 days	32 days	ADA in the LH	Recurrence of EDH (19 mm)	n	n	n	Removal of EDH
4	56/M	Motor aphasia; 13	1 day; 7 days	3 days; 0/4 days; 10 h	4 days; 10 h	ATDA in the LH^§	NC	n	n	n	ASM (add OXZ and increase LVT)
5	78/F	Motor aphasia; 12	8 days	0/50 min	3 days	Low-amplitude ATDA in the LH	NC	NC	n	n	Hydration
6	83/M	Motor aphasia evolving to global aphasia; 13	2 days	0/1 day	1 day	ADA in the LH	NC	NC	n	n	Observation
7	76/M	Global aphasia/right hemianopia; 10	Immediately after trauma	0/5 days	12 days	ADA in all leads^*	Not relevant lesion	Not relevant lesion	n	n	Observation
8	79/F	Decreased fluency/paraphasia; 13	4 days; 9 days; 10 days	0/6 h; 5–12 h; 4 h	6 h; 5–12 h; 4 h	Low-amplitude ADA in the left temporal area	NC	NC	No significant abnormality	No significant abnormality	ASM (FOS) at the first episode; none for the second episode
9	47/F	Motor aphasia evolving to global aphasia; 11	10 days	1 day/4 days	22 days	ADA in the LH	Midline shift (7 mm)	NC	n	n	Craniotomy and removal of SDH/ICH; ASM (add VPA)
10	60/M	Motor aphasia, intermittent; 12t	14 days	0/17 days	17 days	Normal^†	NC	NC	No significant abnormality	n	Observation
11	83/F	Motor aphasia/right arm monoparesis evolving to global aphasia/monoplegia; 10	13 days	1 day/3 days	28–42 days	ADA in the LH	NC	NC	No significant abnormality	Increased CBF in the left temporoparietal lobe	ASM (add OXZ followed by VPA)
12	40/M	Sensory aphasia evolving to global aphasia; 10	2 days	1 day/21 days	23 days	ATDA in the left temporal area	Subtle increase in SDH (3 mm)	NC	No significant abnormality	n	Craniectomy and removal of SDH/EDH Hydration
13	47/M	Global aphasia; 13; 10	8 days; 52 days	0; 0/7 days; 1 day	34 days; 1 day	ATDA in the left temporal area	NC	NC	Diffuse luminal irregularity at both distal ICAs	n	ASM (add OXZ)
14	62/F	Global aphasia; 11	8 days	0/2 days	14–41 days	ATDA in all leads	NC	NC	Multifocal stenosis in the right MCA, both PCAs, and VBA	n	Hydration

Time to peak and time to improving phase were counted from the onset of TND.

Electroencephalography was conducted during the episode of TND except patient 4^§ and 10^†, in whom the recording was performed immediate after TND and between TND symptoms, respectively. Electroencephalography was recorded 12 h after lorazepam injection in patient 7^*.

The modality of vascular imaging was MR angiography in all patients. CT angiography in patient 8 and transfemoral cerebral angiography in patient 14 were additionally performed.

ADA, arrhythmic delta activity; ASM, anti-seizure medication; ATDA, arrhythmic theta to delta activity; CBF, cerebral blood flow; CT, computed tomography; DWI, diffusion-weighted MR imaging; EDH, extradural hemorrhage; FOS, fosphenytoin; GCS, Glasgow Coma Scale; ICA, internal carotid artery; ICH, intracranial hemorrhage; LH, left hemisphere; LVT, levetiracetam; MCA, middle cerebral artery; MRA, magnetic resonance angiography; n, not performed; NC, no significant or relevant change; OXZ, oxcarbazepine; PCA, posterior cerebral artery; PWI, perfusion weighted MR imaging; SDH, subdural hemorrhage; TFCA, transfemoral cerebral angiography; TND, transient neurological dysfunction; VBA, vertebrobasilar artery; VPA, valproate.

All patients were in a stable or improving condition with no (n = 9) or mild neurological dysfunction (n = 4) before the onset of TND except for one patient who initially presented with the index episode (patient 7). Eight patients complained of headache before TND. Systolic blood pressure ranged from 100 to 160 around the beginning of TND.

Twelve patients had SDH in combination with SAH (patients 1, 3, and 13), EDH (patients 1, 3, 6, and 12), and intracerebral hemorrhage (patients 3, 7, 9, and 10) around the symptomatic area, whereas the other two patients (patients 7 and 8) had isolated SAH in the area correlated with TND. The relative risk for TND of patients with SDH over those without SDH was 4.23 (95% confidence interval 0.94–19.04). The lesions were more prominent in the left hemisphere in all patients. Surgery was conducted in seven patients, including craniotomy in six and craniectomy with extra-axial hemorrhage removal in one (patient 13) before index episodes. Prophylactic levetiracetam was started at presentation in all patients.

In all patients, the symptoms of TND included aphasia. Two patients had hemiparesis (patients 3 and 11), and one patient had hemianopia (patient 7) in addition to aphasia. All patients were alert and cooperative. They had neither seizures nor systemic complications around the index episodes. Three patients complained of headache during TND. The headache persisted from initial presentation in two patients (patients 4 and 9), and the onset of headache was unknown in the other patient (patient 7) because he had global aphasia immediately after trauma. The episodes were monophasic in most patients but multiphasic in three patients (patients 4, 8, and 13). The symptoms developed anytime between immediately after trauma and within 14 days after trauma (median 8 days, IQR 2–9 days), which was insidious with progression up to 3 days in eight patients and abrupt in the other six patients. Improvements began from 50 min to 21 days and were gradual in all patients. Symptoms fluctuated in eight patients. The overall duration of index episodes ranged from 4 h to approximately 1 month. Eventually, 13 patients fully recovered from the episodes. One patient (patient 14) had remaining mild comprehension difficulty when evaluated at 2 months after the onset of TND.

Regarding the overall outcome of TBI, 10 patients regained premorbid functional status within three months. The other 4 patients had mild hemiparesis (patient 3), frailty (patient 7), and cognitive dysfunction (patients 13 and 14) at 3 months. Seizures were reported in 2 patients during follow-up. Patient 9 had a focal motor seizure of the face 1 year after TBI, and patient 12 had two bouts of generalized tonic–clonic seizures 78 days after TBI.

Brain CT showed no interval change in 11 patients and progression of TBI in the other 3 patients (patients 3, 9, and 12). DWI was conducted in 12 patients but revealed no additive finding. MRA performed in 7 patients showed no significant abnormality of intracranial arteries except in one patient with severe stenosis of the right middle cerebral artery, which was not on the symptomatic side and already existed before TBI (patient 14). CT angiography or transfemoral cerebral angiography performed additionally in two patients revealed no vasospasm. PWI was performed in two patients, revealing a modest increase in CBF in the symptomatic area in one patient (patient 11) and no significant abnormality in the other patient (patient 8). The symptoms persisted before and after PWI in patient 11, whereas the symptoms fluctuated and disappeared during the acquisition of PWI in patient 8.

Scalp electroencephalography was conducted during the episodes in all patients, except for two (patients 4 and 10) in whom electroencephalography was performed immediate after TND recovery and between TND symptoms, respectively. Electroencephalography was recorded 12 h after lorazepam injection in patient 7. Electroencephalography did not show any epileptiform discharges but revealed persistent arrhythmic mixed slowing in the symptomatic hemisphere in 11 patients, diffuse slowing in 2 patients (patients 7 and 14), and no abnormality in 1 patient (patient 10).

Spectral analysis of electroencephalography was performed in 12 patients except 2 patients (patients 13 and 14) with profuse artifacts for whole recording time. Median frequency at each electrode was lower in the left frontotemporal electrodes (F7, T3, and T5) than in the corresponding electrodes (F8, T4, and T6) of the right hemisphere. Average spectral power of each hemisphere demonstrated relatively stronger power of slow activity in the left hemisphere in most patients (Fig. 4).

FIG. 4.

Topographical representation of median frequency at each electrode and average frequency spectrum of left and right hemisphere. Median frequency was lower in the left frontotemporal area (F7, T3, and T5) than in the right frontotemporal area (F8, T4, and T6), and the slow activity power was relatively stronger in left hemisphere, except patient 10 who showed normal electroencephalography.

Three patients who had interval changes in brain imaging underwent hemorrhage removal or decompressive surgery, which did not abolish the symptoms (patients 3, 9, and 12). No improvement was achieved by loading doses of ASMs, including oxcarbazepine, valproic acid, and fosphenytoin, administered in five patients. Hydration with normal saline was attempted in three patients but was not effective.

Discussion

A small portion of patients with acute TBI experienced TND that could not be explained by any known disease entity. The characteristic features of post-traumatic TND can be summarized as focal neurological dysfunction, slowing of electroencephalography, increased perfusion, and a prolonged but self-limiting clinical course.

Differential diagnosis

Cerebral edema or increased intracranial pressure was suspected to be the cause of neurological deterioration; therefore, decompressive surgery was performed in three patients. However, symptom development was relatively sudden and irrelevant to the structural changes, and symptom resolution was also not attributed to surgery. Nonconvulsive seizure could be a major differential diagnosis for post-traumatic TND. We ruled out nonconvulsive seizures by performing electroencephalography during the symptoms, and no patients with TND experienced a seizure during the peri-TND period. Ischemia was ruled out by the absence of a lesion with diffusion restriction despite symptoms lasting more than 6 h. Vasospasm secondary to traumatic subarachnoid hemorrhage could be considered a cause of TND, but the amount of subarachnoid hemorrhage was too small to cause vasospasm in all patients. Moreover, vascular imaging, including transfemoral cerebral angiography, confirmed the absence of vasospasm and PWI indicated no reduced perfusion in the symptomatic hemisphere.

Shared features with hemiplegic migraine

We noticed some similarities between post-traumatic TND and hemiplegic migraine. Both conditions develop suddenly with the gradual evolution of symptoms, including aphasia or hemiparesis. Electroencephalography during migraine attacks usually shows attenuation or slow activity in the symptomatic hemisphere.⁹ Most reported cases of hemiplegic migraine showed vasodilation or increased perfusion during a migraine aura on vascular or perfusion imaging.¹⁰ Headache was not the prominent symptom in post-traumatic TND, whereas headache is a major symptom in hemiplegic migraine. However, it was difficult to determine the presence, character, and frequency of headache because many patients already had headache associated with TBI before TND, and all patients had aphasia during TND.

Proposal for pathophysiology

We suggest SD as the underlying mechanism of post-traumatic TND. Cortical spreading depression was suggested as a pathophysiology of migraine based on the similarities between the typical pattern of symptom evolution of migraine aura and the propagation of cortical spreading depression triggered in animal brains despite the absence of direct evidence in humans.^7,11 Later, genetic mutations in ion channels in familial hemiplegic migraine supported the link between cortical spreading depression and migraine.⁹ Cortical spreading depression is a consequence of neuronal SD, characterized by the breakdown of ion gradients, which is often observed in an injured brain. CBF increases in response to SD, similar to normal neurovascular coupling to neuronal depolarization, although neuronal signaling is depressed, and inverse neurovascular response can occur depending on tissue conditions.¹² This feature of SD is consistent with the characteristic finding of electroencephalography slowing and increased perfusion during post-traumatic TND observed in this study. TBI itself shows decreased perfusion.^13,14 Typically, seizures accompany ictal electroencephalography hypersynchrony and increased perfusion, whereas transient ischemic attacks present with electroencephalography slowing and decreased perfusion.

TND has also been observed in other neurosurgical conditions. We reported prolonged but spontaneously recovering hemiparesis and/or dysphasia accompanied by hyper-perfusion without any evidence of structural changes, ischemia/vasospasm, or seizures after indirect revascularization surgery for moyamoya disease.¹⁵ SD recorded from electrocorticography has been reported in TBI as well as in spontaneous intracerebral hemorrhage, SAH, and ischemic stroke.^16

–19 Increased CBF during cortical spreading depression has been observed during intracranial monitoring of patients with severe TBI.^20,21 However, clinical correlation with SD could not be assessed in most studies because they were performed in comatose patients in whom craniectomy was needed. A recent study that monitored electroencephalography after chronic SDH evacuation demonstrated a TND occurring time-locked with a series of SD clusters.²² A study that examined SD after SAH reported the occurrence of neurological deficit during or following the clusters of SD. Because they were frequently associated with vasospasm and development of infarction, SD was interpreted as a sentinel sign of ischemia in the study. However, it is notable that a patient had repeated spreading depression accompanied by neurological deficit with neither vasospasm nor infarction, which is analogous to post-traumatic TND.²³

Despite the symptoms being prolonged, most of our patients fully recovered without any sequelae as seen in migraine and TND after concussion,⁴ whereas previous reports of severe TBI and stroke have suggested that the presence of prolonged SD is associated with an unfavorable clinical outcome.^18,23,24 SD may not necessarily indicate poor prognosis, as implicated by the concept of the “SD continuum.”²⁵

Various types of post-traumatic TND have been reported with different terminologies. Several reports have shown that blow-to-head triggered attacks resembling spontaneous hemiplegic migraine episodes experienced by the patients or their relatives. Classical migraine with visual aura precipitated by a head-on collision while playing football has also been reported. Therefore, this was referred to as “trauma-triggered migraine,” “post-traumatic migraine,” or “footballer’s migraine.”^4,5,26 Although postconcussion TND has mostly been described in children and young adults,⁴ chronic SDH presenting with recurrent TND, described as “transient neurological deficit simulating transient ischemic attack (TIA),” has been reported in the elderly. Evacuation surgery was conducted in all 14 patients, and TND persisted in three patients even after surgery in a previous report.⁶ Recently, there have been anecdotal reports of unexplained TND after the evacuation of acute or chronic SDH.^27

–30 In two cases, ASMs were tried but were not effective.^27,30 TND resolved after discontinuation of antihypertensive medications and administration of fluid bolus in one patient who had relatively low blood pressure at the event. They suggested hemodynamic compromise by edema or fluctuating intracranial pressure, naming this phenomenon a “post-subdural hematoma TIA.”²⁸ Seizure undetected on scalp electroencephalography was suggested in a study based on the occurrence of electrographic seizures 2–3 days after TND in three patients.²⁹ Another researcher proposed the term “nonepileptic stereotypical and intermittent symptoms” for TND occurring after SDH.³¹ A recent study presented recurrent hemiparesis in an elderly person who had traumatic SAH without vascular stenosis/spasm or epileptic discharge and proposed to call the phenomenon “spreading focal neurological deficit” as a kind of “symptomatic migraine aura.”³²

In this study, all the patients with TND had SDH, but two patients had SAH without SDH in the symptomatic area. The higher risk of TND in SDH is uncertain because SDH accounts for the majority of TBI admissions. The incidence of TND was reported as 9% of chronic SDH,⁶ 7% among patients operated for any type of SDH,³¹ and 23% among those who underwent surgery for chronic SDH.³³ The frequency of TND was lower in this study. This could be due to strict inclusion criteria for TND, such as exclusion of nonfocal symptoms in this study and different populations.

TND developed with various latent intervals up to 14 days from trauma in this study, whereas it mostly developed within 1 day in postconcussion TND.⁴ The direct impact of trauma could trigger SD presenting TND immediately after trauma, such as “stunning,” suggested in ischemic brain and myocardium.³⁴ In regard to delayed TND, we speculate that delayed cerebral edema³⁵ and inflammation of the cortex underneath the SDH³⁶ predispose the patients to SD.

Aphasia was the predominant symptom, and all TNDs originated from the left hemisphere in this study. Previous reports of TND after SDH also reported that aphasia was the most common symptom of TND,⁶ and the symptom that favored TND rather than seizure³¹ as well as TND occurred more frequently in left hemispheric lesions.⁶ Whether this is due to selective vulnerability of the language cortex and left hemisphere for post-traumatic TND or easier detection of left than right hemispheric symptoms should be tested by prospective study.

Prevention and Treatment

We administered ASMs as a therapeutic trial of possible nonconvulsive seizures and SD. However, TND was not abolished by any ASMs but improved spontaneously. Similarly, there have been no treatments proven to be effective for abolishing migraine aura. Instead, certain medications, such as calcium channel blockers, tricyclic antidepressants, and ASMs, have been reported to prevent migraine attacks. In this study, the preventive effect of each ASM for TND could not be assessed because most patients received levetiracetam before TND. Among sedatives, Ketamine has been reported to have the suppressive effect on SD.^37,38 Therefore, Ketamine is a future candidate for therapeutic application to subside post-traumatic TND and to test the TND-SD linkage hypothesis.

Strengths and Limitations

This study delineates post-traumatic TND, which could be misinterpreted or overlooked, and provides new insight into its pathophysiology, which has been controversial. Awareness of this phenomenon will help clinicians avoid unnecessary interventions and predict outcomes. The limitation of this study was that it was not a prospective study with a standard protocol. Therefore, this study could underestimate the incidence of TND. The history of migraine in patients with TND was not acquired. It is difficult to ensure that increased perfusion is a consistent feature of TND because perfusion imaging was conducted in only a small number of patients, and the natural course of hemodynamics in TBI itself is not well known. To examine the SD hypothesis for the underlying mechanism of TND, further research including electrocorticography is needed.

Conclusion

Post-traumatic TND is not well recognized but is certainly an existing phenomenon. Post-traumatic TND can persist longer than classical paroxysmal disorders. It is characterized by electrical hypoactivity and perfusion increase, reminiscent of neurovascular response observed in SD, which is distinctive from seizures with electrical hypersynchrony and transient ischemic attack with perfusion decrease. TND can be a differential diagnosis of neurological deterioration in TBI. If other causes that need urgent intervention are ruled out, we can adopt the wait-and-see policy for spontaneous resolution. Preventive and abolitionary therapies for TND should be tested in the future.

Footnotes

Authors’ Contributions

S.-Y.L., S.J.L., and J.H.P. contributed to the study conception and design. Acquisition and analysis of data were performed by S.-Y.L., S.J.L., H.S.J., C.O., and S.S.K. The first draft of the article was written by S.-Y.L. and J.H.P. All authors commented on the previous version of the article, and S.J.L. and S.S.K. edited the article.

Author Disclosure Statement

No competing interests exist.

Funding Information

This research was not supported by any funding agency. S.-Y.L. received a grant from the Korea Health Technology R&D Project, funded by the Ministry of Health & Welfare, Republic of Korea (grant number: HI19C0481, HC19C0180).

References

Marincowitz

, Lecky

, Townend

, et al. The risk of deterioration in GCS13–15 patients with traumatic brain injury identified by computed tomography imaging: A systematic review and meta-analysis. J Neurotrauma, 2018; 35(5):703–718; doi: 10.1089/neu.2017.5259

Choudhry

, Prestigiacomo

, Gala

, et al. Delayed neurological deterioration after mild head injury: Cause, temporal course, and outcomes. Neurosurgery, 2013; 73(5):753–760; discussion 760; doi: 10.1227/NEU.0000000000000105

Al-Mufti

, Amuluru

, Changa

, et al. Traumatic brain injury and intracranial hemorrhage–induced cerebral vasospasm: A systematic review. Neurosurg Focus, 2017; 43(5):E14

Haas

, Lourie

. Trauma-triggered migraine: An explanation for common neurological attacks after mild head injury: Review of the literature. J Neurosurg, 1988; 68(2):181–188; doi: 10.3171/jns.1988.68.2.0181

Solomon

. John graham senior clinicians award lecture. posttraumatic migraine. Headache, 1998; 38(10):772–778; doi: 10.1046/j.1526-4610.1998.3810772.x

Moster

, Johnston

, Reinmuth

. Chronic subdural hematoma with transient neurological deficits: A review of 15 cases. Ann Neurol, 1983; 14(5):539–542; doi: 10.1002/ana.410140508

Leao

AAP

. Spreading depression of activity in the cerebral cortex. Journal of Neurophysiology, 1944; 7(6):359–390; doi: 10.1152/jn.1944.7.6.359

Somjen

. Mechanisms of spreading depression and hypoxic spreading depression-like depolarization. Physiol Rev, 2001; 81(3):1065–1096; doi: 10.1152/physrev.2001.81.3.1065

Russell

, Ducros

. Sporadic and familial hemiplegic migraine: Pathophysiological mechanisms, clinical characteristics, diagnosis, and management. Lancet Neurol, 2011; 10(5):457–470; doi: 10.1016/S1474-4422(11)70048-5

10.

Choi

K-H

, Kim

, Lee

S-Y

, et al. Familial hemiplegic migraine with prolonged coma and cerebellar atrophy: CACNA1A T666M mutation in a Korean family. J Korean Med Sci, 2012; 27(9):1124–1127; doi: 10.3346/jkms.2012.27.9.1124

11.

Pietrobon

, Striessnig

. Neurobiology of migraine. Nat Rev Neurosci, 2003; 4(5):386–398; doi: 10.1038/nrn1102

12.

Dreier

. The role of spreading depression, spreading depolarization and spreading ischemia in neurological disease. Nat Med, 2011; 17(4):439–447; doi: 10.1038/nm.2333

13.

Douglas

, Chaudhari

, Zhao

, et al. Perfusion imaging in acute traumatic brain injury. Neuroimaging Clin N Am, 2018; 28(1):55–65; doi: 10.1016/j.nic.2017.09.002

14.

Smith

LGF

, Milliron

, Ho

M-L

, et al. Advanced neuroimaging in traumatic brain injury: An overview. Neurosurg Focus, 2019; 47(6):E17; doi: 10.3171/2019.9.FOCUS19652

15.

Phi

, Lee

, Kang

H-S

, et al. Postoperative transient neurologic dysfunction: A proposal for pathophysiology. J Clin Neurol, 2018; 14(3):393–400; doi: 10.3988/jcn.2018.14.3.393

16.

Fabricius

, Fuhr

, Bhatia

, et al. Cortical spreading depression and peri-infarct depolarization in acutely injured human cerebral cortex. Brain, 2006; 129(Pt 3):778–790; doi: 10.1093/brain/awh716

17.

Dohmen

, Sakowitz

, Fabricius

, et al. Spreading depolarizations occur in human ischemic stroke with high incidence. Ann Neurol, 2008; 63(6):720–728; doi: 10.1002/ana.21390

18.

Hartings

, Bullock

, Okonkwo

, et al. Spreading depolarisations and outcome after traumatic brain injury: A prospective observational study. Lancet Neurol, 2011; 10(12):1058–1064; doi: 10.1016/S1474-4422(11)70243-5

19.

Drenckhahn

, Winkler

MKL

, Major

, et al. Correlates of spreading depolarization in human scalp electroencephalography. Brain, 2012; 135(Pt 3):853–868; doi: 10.1093/brain/aws010

20.

Mayevsky

, Doron

, Manor

, et al. Cortical spreading depression recorded from the human brain using a multiparametric monitoring system. Brain Res, 1996; 740(1-2):268–274; doi: 10.1016/S0006-8993(96)00874-8

21.

Hinzman

, Andaluz

, Shutter

, et al. Inverse neurovascular coupling to cortical spreading depolarizations in severe brain trauma. Brain, 2014; 137(Pt 11):2960–2972; doi: 10.1093/brain/awu241

22.

Mohammad

, Abbas

, Shuttleworth

, et al. Spreading depolarization may represent a novel mechanism for delayed fluctuating neurological deficit after chronic subdural hematoma evacuation. J Neurosurg, 2020; 134(3):1294–1302; doi: 10.3171/2020.1.JNS192914

23.

Dreier

, Woitzik

, Fabricius

, et al. Delayed ischaemic neurological deficits after subarachnoid haemorrhage are associated with clusters of spreading depolarizations. Brain, 2006; 129(Pt 12):3224–3237; doi: 10.1093/brain/awl297

24.

Strong

, Fabricius

, Boutelle

, et al. Spreading and synchronous depressions of cortical activity in acutely injured human brain. Stroke, 2002; 33(12):2738–2743; doi: 10.1161/01.STR.0000043073.69602.09

25.

Dreier

, Reiffurth

. The stroke-migraine depolarization continuum. Neuron, 2015; 86(4):902–922; doi: 10.1016/j.neuron.2015.04.004

26.

Sakas

, Whittaker

, Whitwell

, et al. Syndromes of posttraumatic neurological deterioration in children with no focal lesions revealed by cerebral imaging: Evidence for a trigeminovascular pathophysiology. Neurosurgery, 1997; 41(3):661–667; doi: 10.1097/00006123-199709000-00031

27.

Shah

, Adams

. Episodic aphasia associated with cortical spreading depression after subdural hemorrhage evacuation. Neurohospitalist, 2016; 6(1):NP1–NP4; doi: 10.1177/1941874415583118

28.

Alkhachroum

, Fernandez-Baca Vaca

, Sundararajan

, et al. Post-subdural hematoma transient ischemic attacks: Hypoperfusion mechanism supported by quantitative electroencephalography and transcranial doppler sonography. Stroke, 2017; 48(3):e87–e90; doi: 10.1161/STROKEAHA.117.016388

29.

Driver

, DiRisio

, Mitchell

, et al. Non-electrographic seizures due to subdural hematoma: A case series and review of the literature. Neurocrit Care, 2019; 30(1):16–21; doi: 10.1007/s12028-018-0503-2

30.

Jain

, Remley

, Mohan

, et al. Nonepileptic, stereotypical, and intermittent symptoms after subdural hematoma evacuation. Cureus, 2021; 13(9):e18361; doi: 10.7759/cureus.18361

31.

Levesque

, Iorio-Morin

, Bocti

, et al. Nonepileptic, stereotypical, and intermittent symptoms (NESIS) in patients with subdural hematoma: Proposal for a new clinical entity with therapeutic and prognostic implications. Neurosurgery, 2020; 87(1):96–103; doi: 10.1093/neuros/nyz355

32.

Major

, Huo

, Lemale

, et al. Direct electrophysiological evidence that spreading depolarization-induced spreading depression is the pathophysiological correlate of the migraine aura and a review of the spreading depolarization continuum of acute neuronal mass injury. Geroscience, 2020; 42(1):57–80; doi: 10.1007/s11357-019-00142-7

33.

Meadows

, Davis

, Mohammad

, et al. Spreading depolarization after chronic subdural hematoma evacuation: Associated clinical risk factors and influence on clinical outcome. Neurocrit Care, 2021; 35(Suppl 2):105–111; doi: 10.1007/s12028-021-01339-5

34.

Alexandrov

, Hall

, Labiche

, et al. Ischemic stunning of the brain: Early recanalization without immediate clinical improvement in acute ischemic stroke. Stroke, 2004; 35(2):449–452; doi: 10.1161/01.STR.0000113737.58014.B4

35.

Barzó

, Marmarou

, Fatouros

, et al. Biphasic pathophysiological response of vasogenic and cellular edema in traumatic brain swelling. Acta Neurochir Suppl, 1997; 70:119–122; doi: 10.1007/978-3-7091-6837-0_37

36.

Squier

, Mack

, Green

, et al. The pathophysiology of brain swelling associated with subdural hemorrhage: The role of the trigeminovascular system. Childs Nerv Syst, 2012; 28(12):2005–2015; doi: 10.1007/s00381-012-1870-1

37.

Carlson

, Abbas

, Alunday

, et al. Spreading depolarization in acute brain injury inhibited by ketamine: A prospective, randomized, multiple crossover trial. J Neurosurg, 2019; 130(5):1–7; doi: 10.3171/2017.12.JNS171665

38.

Hertle

, Dreier

, Woitzik

, et al. Cooperative Study of Brain Injury Depolarizations (COSBID). Effect of analgesics and sedatives on the occurrence of spreading depolarizations accompanying acute brain injury. Brain, 2012; 135(Pt 8):2390–2398; doi: 10.1093/brain/aws152