Copyright # Munksgaard 2000
Acta Neurol Scand 2000: 102: 275±283 Printed in UK. All rights reserved
ACTA NEUROLOGICA SCANDINAVICA ISSN 0001-6314
Review article
Transient global amnesia: a review emphasizing pathogenic aspects Pantoni L, Lamassa M, Inzitari D. Transient global amnesia: a review emphasizing pathogenic aspects. Acta Neurol Scand 2000: 102: 275±283. # Munksgaard 2000. The transitory memory disturbance known as transient global amnesia (TGA) remains an enigma from a pathogenic point of view. In spite of its typical benign prognosis, TGA is a frightening experience for patients and their relatives. Moreover, a TGA episode usually leads to extensive investigation of patients in search of organic alterations that might be responsible for the event. Finally, TGA generates queries about therapeutic choices. In this review, we critically re-evaluate the evidence in support of and against the three main pathogenic hypotheses (i.e. ischemia, seizure discharge, and migraine), and we conclude that none of these appears completely convincing. Given the good prognosis and the lack of association with organic and instrumental abnormalities, we advance the hypothesis that TGA may be related to psychological disturbances causing transient alteration in brain metabolism and, consequently, amnesia. Our conclusion has relevant consequences in the evaluation of patients with TGA.
Clinical features and diagnostic criteria
Fisher & Adams in 1958 (1) and then 6 years later, in a seminal paper reporting on 17 cases of transient disturbance of memory (2), were the ®rst to use the name transient global amnesia (TGA) and to provide an overview about its possible causes. In the same period, other authors described similar cases (3±7). TGA is a disturbance characterized by a sudden inability to acquire new information (a de®cit of anterograde memory), usually lasting a few hours and not accompanied by any other focal neurological signs or symptoms. During the spell, long-term anterograde memory of verbal and non-verbal material is markedly impaired. The patient often asks the same questions repetitively to nearby persons and appears perplexed, disoriented in terms of time and sometimes of location but never looses self-awareness. Immediate memory (repetition of a sentence, digit-span) and the ability to
L. Pantoni, M. Lamassa, D. Inzitari Department of Neurological and Psychiatric Sciences, University of Florence, Florence, Italy
Key words: transient global amnesia; pathogenesis; cerebral ischemia; migraine; epilepsy; panic attack; agoraphobia; spreading depression Leonardo Pantoni, Department of Neurological and Psychiatric Sciences, University of Florence, Viale Morgagni 85, 50134 Firenze, Italy Tel.: 39 055 4277 995 Fax: 39 055 4298 461 e-mail:
[email protected]®.it Accepted for publication July 4, 2000
perform some complex activities are preserved. In many instances the subject also presents a de®cit in retrograde memory (8, 9). Within a few hours, seldom more than 12, his memory de®cit resolves and the patient retains a memory gap for the period covered by the attack. The diagnosis of TGA relies on clinical basis. The most widely used criteria are those proposed by Caplan (8) and then modi®ed and validated by Hodges & Warlow (10).
Epidemiology
Speci®c epidemiological studies are scanty. In Rochester, Minnesota, USA, the annual incidence was reported to be 5.2/100,000/year (11), whereas, in a study performed in Finland (12), higher ®gures were reported (10 and 32/100,000/year for the general population, and for patients over 50 years of age, respectively). Hodges (9), evaluating the 275
Pantoni et al. Oxfordshire Community Stroke Project data (13), found that the annual incidence was 3/100,000/year. Lauria et al. (14), in a prospective study carried out in Belluno, Italy, con®rmed the higher Finnish frequency and underlined that TGA is not a rare clinical event. The peak of incidence of TGA is in the 7th decade (8, 9, 11). No obvious gender difference in the incidence of TGA is reported. The percentage of TGA recurrences varies largely across studies (from 2.9 to 42%) (11, 14±25). This broad variation could depend on the duration of the follow-up. Based on published series, Hinge & Jensen (26) reported a 2.5% annual risk for a TGA recurrence.
Pathogenic hypotheses
In the past, a number of amnesic spells occurring in patients with de®nite brain organic diseases (27±30) have been reported as TGAs. As outlined by Hodges (9), many of these cases were not completely consistent with current diagnostic criteria. Moreover, the ®nding of some organic abnormalities (e.g. an otherwise silent meningioma during a computed tomography evaluation of a TGA patient) does not necessarily mean that the lesion is etiologically correlated with the clinical picture. In addition to these anecdotal cases, three main hypotheses on the pathogenesis of TGA have been proposed.
TGA as a transient ischemic attack
According to some authors, TGA would result from transient ischemia caused by thrombo-embolic or hemodynamic mechanisms in brain regions deputed to memory (25, 31, for a review see 32, 33). This hypothesis was initially advanced on the basis of non-controlled studies that showed, among TGA patients, an apparently high prevalence of vascular risk factors (24, for a review see 9, 34) and a high rate of subsequent stroke (35). Thus, TGA was considered a sort of amnesic transient ischemic attack (TIA) (36, 37), also in keeping with reports of amnesic strokes (38±41). Recent controlled studies have demonstrated instead that patients with a wellde®ned diagnosis of TGA have the same prevalence of vascular risk factors as the general population (16, 17, 19, 42) and a better prognosis than TIA patients (11, 16, 17, 19, 20, 22, 42). Moreover, among TGA patients, laboratory investigations (cranial computed tomography, neck vessels ultrasounds, and cerebral angiography) usually show only slight abnormalities, mainly consistent with aging (20, 37, 43). 276
TGA as an epileptic phenomenon
A 2nd hypothesis interprets TGA as an epileptic phenomenon involving the hippocampal±diencephalic system (44, for a review see 45). This theory is based on some similarities between TGA episodes and epileptic attacks selectively involving memory (the so-called ``pure amnestic'' seizures) (46±48). The results of EEG registrations performed during (49±51) and after (4, 18, 24, 51, 52) pure TGA episodes which, with few exceptions (53), do not show any epileptic activity stand against this hypothesis. Furthermore, the absence of epileptic clinical features (automatisms, altered consciousness, etc.) during TGA, the long duration and the low rate of recurrence of the episodes, make an epileptic origin of TGA unlikely. However, some authors have reported that particularly brief (lasting
