Diffusion-Weighted MRI in Transient Global Amnesia Precipitated by Cerebral Angiography
Background Transient global amnesia is a well-described complication of cerebral angiography. Speculation about the pathophysiology exists but is as yet unsubstantiated. Diffusion-weighted MRI is a new imaging technique that is very sensitive in detecting acute ischemia. Its use in the evaluation of transient amnesia precipitated by cerebral angiography has not previously been reported.
Case Description A 44-year-old man underwent posterior circulation cerebral angiography for the investigation of episodic vertigo. Shortly after completion of the procedure, he was noted to have symptoms of transient global amnesia. Diffusion-weighted MRI at 6 and 44 hours after the procedure demonstrated increased signal in the right hippocampus and other areas within the posterior circulation bilaterally consistent with ischemia from emboli. Abnormalities on conventional MRI images performed at the same time points were noted only in retrospect. A follow-up MRI at 2 months was normal.
Conclusions Ischemia from cerebral emboli may cause transient global amnesia precipitated by cerebral angiography. Diffusion-weighted MRI may be useful in defining the pathophysiology.
Transient global amnesia refers to a syndrome of temporary anterograde and recent retrograde memory loss with preservation of alertness, attention, and self-identity, without other significant accompanying neurological symptoms. Although not the first to report TGA, Fisher and Adams1 are credited with naming the syndrome and monographing it in detail. In addition, TGA is a well-recognized complication of cerebral angiography when either ionic or nonionic contrast agents are used.2 The precise incidence after cerebral angiography is uncertain but is estimated at 0.8%3 with ionic contrast agents. Thus far, only 14 such patients have been reported in the literature after nonionic contrast use,4 consistent with the speculation that the incidence is reduced with the use of nonionic agents and modern catheters.2 Neuroimaging studies in patients who develop TGA after angiography have failed to demonstrate acute lesions.
DWI is a novel MRI technique that is very sensitive in detecting acute ischemia minutes after onset.5 Minimal information is available about its use in the evaluation of acute amnesia.
We describe a patient who developed transient amnesia, clinically identical to TGA, after cerebral angiography of the posterior circulation. Increased signal within the right hippocampus was identified on echo-planar DWI-MRI 6 and 44 hours after symptom onset and was associated with a corresponding reduction in the ADC consistent with acute ischemia. Other less conspicuous areas of increased signal were noted and suggested a more diffuse and bilateral PCA territory ischemic insult. Abnormalities on conventional MRI sequences were only appreciated in retrospect after comparison with the diffusion images. Detailed follow-up MRI with DWI-MRI at 60 days failed to show changes consistent with cerebral infarction.
A 44-year-old right-handed man underwent cerebral angiography to evaluate recurrent isolated vertigo. In the preceding 6 months he had several episodes of vertigo lasting days associated with nausea and vomiting. There were no other neurological symptoms. There was a history of hypercholesterolemia and infrequent migraine without aura but no other stroke risk factors. His last migraine was 4 months before the angiogram. His past history was otherwise unremarkable. His general and neurological examinations were normal. An MRI revealed a dolichoectatic right vertebral artery causing an impression on the right anterolateral aspect of the cervicomedullary junction (not shown).
Cerebral angiography was performed with the use of iohexol, a nonionic contrast agent. The right and left vertebral arteries were selectively studied. The procedure went smoothly. The patient’s symptoms of amnesia were first noted in the recovery room, 10 minutes after angiography. The Stroke Service was subsequently notified, and the patient was assessed 2 hours after the onset of his symptoms.
On examination the patient was alert and orientated to person but to neither place nor time. He repeated the same questions over and over. He could not remember having an angiogram but did know he was scheduled to have one. He had normal attention and immediate recall, but anterograde verbal and nonverbal memory were impaired. Retrograde memory for recent events was similarly affected, but remote memory was intact. The remainder of the neurological examination was normal. Over the next few hours he improved, although he still had mild anterograde and recent retrograde amnesia. He remained amnestic for the angiogram. The following morning, approximately 20 hours after the angiogram, he had normal anterograde memory but could not remember the events of the preceding day, including the angiogram. At follow-up 2 months after the event, this memory gap persisted without other memory problems.
Initial MRI imaging 6 hours after angiography was performed which included sagittal T1, axial FSE T2 and proton-density, FLAIR, and multishot isotropic echo-planar diffusion sequences. On the initial MRI, a lesion in the right hippocampus was seen on the isotropic echo-planar DWI sequence with a corresponding reduction in the ADC consistent with acute infarction (Fig 1⇓). Several minute areas of increased signal within both occipital lobes were also seen on the DWI-MRI suggestive of ischemia but were too small for ADC calculations. The left hippocampus and right thalamus were normal. T1, FSE T2, FSE proton-density, and FLAIR images were initially interpreted as normal. Retrospective analysis at a computer workstation allowed precise direct comparison of the diffusion-weighted images with the other sequences. When a region of interest was outlined on the DWI-MRI and transferred to the T2 images, a subtle corresponding area of hyperintensity within the right hippocampus was appreciated (Fig 1⇓). A repeat MRI 44 hours after angiography showed the conspicuous right hippocampal abnormality on the diffusion-weighted images, but once again the T2 abnormalities were only identified by retrospective analysis. On the DWI sequences, the prior minute occipital abnormalities were more easily identified, and suspicious changes were also seen in the right thalamus and left temporal lobe (Fig 2⇓). A follow-up MRI with DWI-MRI performed 2 months later was normal (Fig 3⇓).
Our patient satisfies the criteria of Caplan6 and Hodges7 for the diagnosis of TGA. In patients with nonangiographic-related TGA, suggested etiologies include ischemic insult, migrainous phenomenon including spreading depression, or epileptic seizure.8 After cerebral angiography, postulated mechanisms for TGA include ischemia (from catheter tip, atherosclerotic plaque, or contrast particulate embolus or vasospasm) or direct neurotoxic effects of the contrast.2 9 In our patient the DWI-MRI findings of minute multifocal bilateral PCA territory lesions in combination with the clinical setting are most suggestive of ischemia from emboli. Ischemia from vasospasm would not likely produce the pattern of multifocal changes observed on the DWI-MRI. Temporal lobe seizure induced by ischemia is plausible but unlikely since transient epileptic amnesia manifests in a briefer (<1 hour) and clinically different fashion.10 The presence of MRI abnormalities 44 hours after angiography, at least 28 hours after the resolution of the patient’s TGA, argues against spreading depression in isolation.
Although ischemia is frequently mentioned as a potential cause of TGA after angiography, to our knowledge our report is the first to substantiate this hypothesis. Prior reports have utilized less sensitive imaging techniques. Without DWI-MRI, we would have misinterpreted the conventional MRI sequences as normal.
Additional lesions were present on the 44-hour DWI-MRI that were not seen on the initial DWI-MRI. This is consistent with the observation that volume and intensity of DWI lesions may increase in the first 2 and 7 days, respectively, after stroke.11 12 Volume averaging may have also contributed in this case. The patient’s TGA had resolved before the 44-hour MRI, making recurrent or ongoing ischemia doubtful.
We are uncertain why the right hippocampal infarction was not seen on the follow-up MRI at 60 days. Both axial and coronal FLAIR sequences, without slice gaps, in addition to other conventional and diffusion sequences were obtained. Despite these techniques, an infarction was probably too small to visualize at this time point. Another possibility, although less likely, is that the initial MRI abnormalities were reversible, with the reversible conventional MRI findings representing vasogenic rather than cytotoxic edema. Ordinarily, vasogenic edema is bright on the ADC map,13 opposite to the findings seen in our patient. Furthermore, although reversible DWI-MRI changes have been documented in animals with the use of focal ischemic models,5 14 they have only rarely been demonstrated in humans.15 Therefore, we believe the most likely explanation for the absence of infarction on the 60-day scan is that small brain infarctions occurred but were too small to be seen.
Our patient had unambiguous right hippocampal and bilateral occipital changes on two DWI-MRIs with equivocal right thalamic and left temporal involvement. Traditional teaching and a majority of the literature suggest that short-term retrograde associated with anterograde memory disorder in humans is caused by bilateral temporal lobe/temporal lobe connection (including the thalamus) dysfunction,16 while others have implicated the hippocampi specifically.17 On the other hand, Ott and Saver18 reported six cases of unilateral stroke (including one patient with TGA) and a review of the literature, demonstrating that unilateral mesial temporal or thalamic stroke (left more common than right) may present with isolated amnesia. Unilateral retrosplenial lesions (Fig 2C⇑) may also cause amnesia.19 In addition, single-photon emission CT and positron emission tomography studies have demonstrated unilateral abnormalities in patients with TGA.20 On this basis, transient unilateral right hippocampal, retrosplenial, and thalamic dysfunction is feasible; however, the majority of the literature and the apparent bilateral PCA territory involvement argue for bilateral dysfunction of memory structures in our case.
It is unlikely that our findings extrapolate to TGA in general. It is believed that TGA is a benign condition.6 Detailed reviews have failed to demonstrate significant vascular risk factors and subsequent increased risk of stroke in these patients when considered as a group.21 However, Bruening and colleagues22 report in abstract form the results of DWI-MRI in patients with nonangiographic-related TGA. Of 10 patients studied, 7 had reversible abnormalities in the hippocampus (4 left sided, 3 bilateral) that they believed were consistent with spreading depression.
From our observations, we conclude that ischemia due to cerebral emboli is a cause of TGA after cerebral angiography. Second, DWI-MRI may be useful in defining the pathophysiology not only in this group but also in those patients with cryptogenic TGA.
Selected Abbreviations and Acronyms
|ADC||=||apparent diffusion coefficient|
|FLAIR||=||fluid attenuation inversion recovery|
|FOV||=||field of view|
|NEX||=||number of excitations|
|PCA||=||posterior cerebral artery|
|TGA||=||transient global amnesia|
This study was supported in part by National Institutes of Health grant NS34866-01A1. The authors would like to express their gratitude to G.W. Albers, MD, for critical review of the manuscript and to M.P. Marks, MD, for analysis of the imaging and helpful suggestions.
- Received July 2, 1997.
- Revision received August 6, 1997.
- Accepted August 6, 1997.
- Copyright © 1997 by American Heart Association
Fisher CM, Adams RD. Transient global amnesia. Acta Neurol Scand. 1964;40(suppl 9):1-83.
Pexman JHW, Coates RK. Amnesia after femorocerebral angiography. AJNR Am J Neuroradiol. 1983;4:979-983.
Hasegawa Y, Fisher M, Latour LL, Dardzinski BJ, Sotak CH. MRI diffusion mapping of the reversible and irreversible ischemic injury in focal brain ischemia. Neurology. 1994;44:1484-1490.
Caplan LR. Transient global amnesia: criteria and classification. Neurology. 1986;36:441.
Hodges JR, Warlow CP. The aetiology of transient global amnesia: a case-control study of 114 cases with prospective follow-up. Brain. 1990;113:639-657.
Jackson A, Stewart G, Wood A, Gillespie JE. Transient global amnesia and cortical blindness after vertebral angiography: further evidence for the role of arterial spasm. AJNR Am J Neuroradiol. 1995;16:955-959.
Kapur N. Transient epileptic amnesia: a clinical update and a reformulation. J Neurol Neurosurg Psychiatry. 1993;56:1184-1190.
Schaefer PW, Buonanno FS, Gonzalez RG, Schwamm LH. Diffusion-weighted imaging discriminates between cytotoxic and vasogenic edema in a patient with eclampsia. Stroke. 1997;28:1082-1085.
Horel JA. The neuroanatomy of amnesia: a critique of the hippocampal memory hypothesis. Brain. 1978;101:403-445.
Ott BR, Saver JL. Unilateral amnestic stroke: six new cases and a review of the literature. Stroke. 1993;24:1033-1042.
Valenstein E, Bowers D, Verfaellie M, Heilman KM, Day A, Watson RT. Retrosplenial amnesia. Brain. 1987;110:1631-1646.
Baron JC, Petit-Taboue MC, Le Doze F, Desgranges B, Ravenel N, Marchal G. Right frontal cortex hypometabolism in transient global amnesia: a PET study. Brain. 1994;117:545-552.
Zorzon M, Antonutti L, Mase G, Biasutti E, Vitrani B, Cazzato G. Transient global amnesia and transient ischemic attack: natural history, vascular risk factors, and associated conditions. Stroke. 1995;26:1536-1542.
Bruening R, Strupp M, Wu RH, Deimling M, Heuck A, Reiser M Brandt TH. Transient global amnesia: evidence of decreased interstitial volume in diffusion-weighted MRI. In: Proceedings of the 5th annual meeting of the International Society for Magnetic Resonance in Medicine; April 12-18, 1997; Vancouver, Canada. Abstract 271.