Skip main navigation

Ischemic Amnesia

Causes and Outcome
Originally published 2017;48:2270–2273


Background and Purpose—

We aimed to describe the frequency and characteristics of acute ischemic stroke and transient ischemic attacks presenting predominantly with amnesia (ischemic amnesia) and to identify clinical clues for differentiating them from transient global amnesia (TGA).


We retrospectively analyzed and described all patients presenting with diffusion-weighted imaging magnetic resonance imaging–confirmed acute ischemic stroke/transient ischemic attacks with antero- and retrograde amnesia as the main symptom over a 13.5-year period. We also compared their clinical features and stroke mechanisms with 3804 acute ischemic stroke from our ischemic stroke registry.


Thirteen ischemic amnesia patients were identified, representing 0.2% of all patients with acute ischemic stroke/transient ischemic attack. In 69% of ischemic amnesia cases, amnesia was transient with a median duration of 5 hours. Ischemia was not considered in 39% of cases. Fifty-four percent of cases were clinically difficult to distinguish from TGA, including 15% who were indistinguishable from TGA. 1.2% of all presumed TGA patients at our center were later found to have ischemic amnesia. Amnesic strokes were more often cardioembolic, multiterritorial, and typically involved the posterior circulation and limbic system. Clinical clues were minor focal neurological signs, higher age, more risk factors, and stroke favoring circumstances. Although all patients were independent at 3 months, 31% had persistent memory problems.


Amnesia as the main symptom of acute ischemic cerebral events is rare, mostly transient, and easily mistaken for TGA. Although clinical clues are often present, the threshold for performing diffusion-weighted imaging in acute amnesia should be low.


Transient and persistent amnesia may be the sole or main manifestation of stroke.1 Other diseases may cause acute amnesia, including transient global amnesia (TGA), epileptic seizures (transient epileptic amnesia), migraine with aura (migrainous amnesia), limbic encephalitis, and dissociative amnesia.2,3 Based on clinical criteria alone, it may be difficult or impossible to differentiate causes of acute amnesia, and additional examinations are often required. Also, risk of subsequent diseases varies significantly between different etiologies: transient epileptic amnesia may have higher risk of developing epilepsy and ischemic amnesia of subsequent stroke.

We aimed to describe the clinical characteristics of consecutive patients with diffusion-weighted imaging (DWI)–confirmed ischemic amnesia (stroke or DWI-positive transient ischemic attack [TIA]) presenting at the stroke center of a single institution over a prolonged time period.


The retrospective analysis period extended from January 2002 to June 2015 (13.5 years). One the authors (PM) maintained a prospective list of all acute ischemic stroke and TIA patients presenting with acute amnesia. These patients were registered in the hospital stroke center databases, that is, either the Acute Stroke Registry and Analysis of Lausanne (ASTRAL)4 or the parallel ASTRAL-E registry of patients admitted with a TIA or with a stroke arriving >24 hours after onset.

Patients were included in the current analysis if their predominant presenting sign was acute onset amnesia manifesting as anterograde and retrograde amnesia, with the other higher cortical functions being intact on clinical bedside testing, and if an acute ischemic lesion was identified on DWI-magnetic resonance imaging (MRI). The amnesia had to be witnessed by bystanders or confirmed on bedside tests of mental status by physicians.

The diagnosis of ischemic amnesia had to be confirmed by DWI-MRI. The recommended imaging work-up of patients with acute amnesia in our institution consists of a 1.5 or 3 Tesla MRI with DWI if the history, symptoms, or signs are atypical for TGA (red flags); therefore, the decision to perform an MRI was up to the treating neurologist. Patients not undergoing a DWI and patients without a TGA-compatible DWI lesion were diagnosed as TGA if the classical criteria were fulfilled.5

Patients with transient amnesia with clinical diagnostic criteria of TGA and uni- or bilateral isolated punctiform DWI lesion in the hippocampus were considered TGA.3,6

We calculated the frequency of ischemic amnesia in all patients ressembling TGAs presenting acutely to our institution (emergency room), using the electronic hospital archives for identification of TGAs according to the International Classification of Diseases Tenth Revision classification over the entire observation period. Given that our electronic hospital data system only allows anonymous extraction of retrospective patient data, we were unable to determine how many of the TGA patients had a subacute DWI.

We also compared clinical features and stroke mechanisms of our patients with all consecutive acute ischemic stroke from ASTRAL from 2003 to 2013.

The ethics commission for research on humans of the Canton of Vaud approved the scientific use of ASTRAL and ASTRAL-E.


Over the 13.5-year observation period, we identified 20 acute ischemic stroke patients presenting predominantly with amnesic symptoms; of these, 7 patients were excluded because of presence of other focal symptoms, loss of consciousness, or unwitnessed transient amnesia.

In 5 of the 13 identified patients, ischemia was not considered initially. Two patients were considered certain TGAs before imaging because they fulfilled all clinical TGA criteria.5

Compared with the 164 patients with a discharge diagnosis of TGA, the underdiagnosis rate of ischemic amnesia was 1.2% (2/166). In another 3.0% (5/169) of patients, the stroke would not have been diagnosed without DWI imaging. The true rate of underdiagnosis of ischemic amnesia in our TGA patients is unknown because an estimated <25% of clinical TGA underwent DWI imaging.

Clinically, all patients had sudden onset of amnesic symptoms, and repetitive questioning was present in 2. Pure amnesic symptoms or signs were present in 7 patients and 4 had added minor cognitive signs (executive deficits and mild anomia). Four patients had other minor neurological signs on detailed neurological examination. Amnesia was transient in 69% of patients (9/13), with a median duration of 5 hours (range of 2 minutes to 11 days). It lasted <24 hours in 54% (7) of patients (TIA). Triggering psychological or physical factors were absent, but 4 patients had circumstances favoring stroke, such as coronarography.

On acute or subacute MRI, at least 1 ischemic DWI lesion was found to be localized in the Papez’ circuit in 9 patients and in neighboring structures in 2 others (see the Figure). Two patients without visible lesions in memory structures (Nos. 7 and 11) had multiple small posterior and anterior circulation lesions. Only 1 patient (No. 13) had an isolated, small hippocampal lesion on DWI 20 hours after onset. This lesion was brighter and more linear than the usual punctuate TGA lesions (Figure, image 13a) and showed Gadolinium uptake on repeat MRI at 8 days (Figure, image 13b), leading to the diagnosis of ischemic amnesia.


Figure. Selected diffusion-weighted imaging (DWI) magnetic resonance imaging (MRI) for all patients with amnesic ischemia, showing the main ischemic lesion(s). Some patients had additional lesions on other slices; not all shown lesions were considered sufficient to explain the amnesia. 10a and 10b, Two different slices of the DWI MRI of patient No 10, showing several lesions potentially causing transient amnesia. 12a, Acute computed tomographic (CT) perfusion, and 12b, DWI MRI of patient No 12. 13a, DWI MRI and 13b, acute apparent diffusion coefficient of patient No 13.

The posterior circulation was affected in 12 patients (92%). Four of them had simultaneous anterior circulation lesions, compatible with the high rate of cardioembolic strokes (7/13). The ischemic amnesia strokes were more often cardioembolic (53.8% versus 28.9%), multiterritorial (30.7% versus 6.9%), and involving the posterior circulation (92.3% versus 27.8%) when compared with the nonamnesic acute ischemic stroke; their initial stroke diagnosis was also more often uncertain (76.9% versus 2.4%).

During the 12 months of follow-up, all ischemic amnesia patients were functionally independent. Because of the transient nature of symptoms in most, only 7 patients had formal neuropsychological assessment in the subacute phase. In 4 patients, some amnesia persisted on long-term follow-up, documented by repeat neuropsychological testing in 3 patients and 5-word bedside testing in 1 patient.


During a 13.5-year observation period in a single stroke center, ischemic amnesia was rare, and the correct diagnosis was initially missed in approximately half of the patients. The most frequent erroneous diagnosis was TGA from which it was clinically indistinguishable in 2 patients. Only 1.2% of clinical TGAs were later shown to be ischemic, although this proportion may be an underestimation in our series given the absence of routine DWI in our TGA patients.

In TGA, a subacute punctuate DWI lesion is often seen in the hippocampus,3,7 compatible with a transient perturbation of the hippocampal circuits.8 On the other hand, there are reports of isolated ischemic amnesia affecting the hippocampus.1,9,10 In our series, the ischemic hippocampal lesions we observed were larger (see the Figure) and often associated with other ischemic lesions.6,11 Among the 3 patients with thalamic strokes, only 1 had an isolated anterior lesion, well known for its participation in the amnesic pathways. The other 2 had additional damage in the limbic system. Recently, a study of thalamic strokes had also implicated a mediodorsal nuclei-associated damage to the mamillothalamic tract.12

Ischemic amnesias mostly affected the posterior circulation6,9,13 or multiple territories and were of cardioembolic origin. Involvement of limbic or associated structures was present in 84.6% of our patients on DWI13; it is likely that comparable lesions were not seen in the other 2 patients because of limited imaging resolution or a too short-lived ischemia.

Most amnesias in our patients were transient. When we compare our patients to a published series of TGA patients,14,15 ischemic amnesia seems to occur at higher age in patients with more cardiovascular risk factors and in patients with a known embolic stroke source. A low threshold for performing DWI MRI seems, therefore, reasonable in acutely amnesic patients, in particular if red flags are present clinically, such as unusual duration of amnesia, minor cognitive or neurological signs, or high vascular risk factor load.

Given the rapid hospital arrival of most of our patients and the potential long-term consequences of amnesia, thrombolysis within the usual time windows should be considered for patients with proven ischemic amnesia.

Strengths of our study are the consecutive and systematic collection of relevant patients over a long time period and a well characterized comparison group.4 Limitations are the absence of systematic high-resolution DWI studies in the subacute phase of most of our TGA patients. Also, formal neuropsychological testing was not performed in many of our patients because of the transient nature of their amnesia.

In conclusion, amnesia as the main and presenting symptom of stroke and TIA is infrequent and mostly transient. As it may be mistaken for TGA, physicians should not refrain from performing DWI imaging in acutely amnesic patients, in particular in the presence of red flags.


We thank Prof J. Ferro for helpful comments for the article, and Mrs J. Chevalier Parisod for help with identifying TGA patients.


Correspondence to Patrik Michel, MD, Neurology Service, Stroke Center, Lausanne University Hospital, Bugnon 46, CH 1011 Lausanne, Switzerland. E-mail


  • 1. Ott BR, Saver JL. Unilateral amnesic stroke. Six new cases and a review of the literature.Stroke. 1993; 24:1033–1042.LinkGoogle Scholar
  • 2. Croft PB, Heathfield KW, Swash M. Differential diagnosis of transient amnesia.Br Med J. 1973; 4:593–596.CrossrefMedlineGoogle Scholar
  • 3. Bartsch T, Butler C. Transient amnesic syndromes.Nat Rev Neurol. 2013; 9:86–97. doi: 10.1038/nrneurol.2012.264.CrossrefMedlineGoogle Scholar
  • 4. Michel P, Odier C, Rutgers M, Reichhart M, Maeder P, Meuli R, et al. The Acute STroke Registry and Analysis of Lausanne (ASTRAL): design and baseline analysis of an ischemic stroke registry including acute multimodal imaging.Stroke. 2010; 41:2491–2498. doi: 10.1161/STROKEAHA.110.596189.LinkGoogle Scholar
  • 5. Hodges W. Syndromes of transient amnesia: Towards a classification. A study of 153 cases.J Neurol Neursurg Psych. 1990; 53:834–843CrossrefMedlineGoogle Scholar
  • 6. Förster A, Griebe M, Gass A, Kern R, Hennerici MG, Szabo K. Diffusion-weighted imaging for the differential diagnosis of disorders affecting the hippocampus.Cerebrovasc Dis. 2012; 33:104–115. doi: 10.1159/000332036.CrossrefMedlineGoogle Scholar
  • 7. Sedlaczek O, Hirsch JG, Grips E, Peters CN, Gass A, Wöhrle J, et al. Detection of delayed focal MR changes in the lateral hippocampus in transient global amnesia.Neurology. 2004; 62:2165–2170.CrossrefMedlineGoogle Scholar
  • 8. Peer M, Nitzan M, Goldberg I, Katz J, Gomori JM, Ben-Hur T, et al. Reversible functional connectivity disturbances during transient global amnesia.Ann Neurol. 2014; 75:634–643. doi: 10.1002/ana.24137.CrossrefMedlineGoogle Scholar
  • 9. Carota A, Lysandropoulos AP, Calabrese P. Pure left hippocampal stroke: a transient global amnesia-plus syndrome.J Neurol. 2012; 259:989–992. doi: 10.1007/s00415-011-6296-2.CrossrefMedlineGoogle Scholar
  • 10. Szabo K, Förster A, Jäger T, Kern R, Griebe M, Hennerici MG, et al. Hippocampal lesion patterns in acute posterior cerebral artery stroke: clinical and MRI findings.Stroke. 2009; 40:2042–2045. doi: 10.1161/STROKEAHA.108.536144.LinkGoogle Scholar
  • 11. Förster A, Al-Zghloul M, Wenz H, Böhme J, Groden C, Neumaier-Probst E. Isolated punctuate hippocampal infarction and transient global amnesia are indistinguishable by means of MRI.Int J Stroke. 2017; 12:292–296. doi: 10.1177/1747493016676613.CrossrefMedlineGoogle Scholar
  • 12. Danet L, Barbeau EJ, Eustache P, Planton M, Raposo N, Sibon I, et al. Thalamic amnesia after infarct: The role of the mammillothalamic tract and mediodorsal nucleus.Neurology. 2015; 85:2107–2115. doi: 10.1212/WNL.0000000000002226.CrossrefMedlineGoogle Scholar
  • 13. Ferro JM. Hyperacute cognitive stroke syndromes.J Neurol. 2001; 248:841–849.CrossrefMedlineGoogle Scholar
  • 14. Melo TP, Ferro JM, Ferro H. Transient global amnesia. A case control study.Brain. 1992; 115 pt 1:261–270.CrossrefMedlineGoogle Scholar
  • 15. Mangla A, Navi BB, Layton K, Kamel H. Transient global amnesia and the risk of ischemic stroke.Stroke. 2014; 45:389–393. doi: 10.1161/STROKEAHA.113.003916.LinkGoogle Scholar


eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.