(Stroke. 1995;26:1536-1542.)
© 1995 American Heart Association, Inc.
Articles |
From the Department of Neurology, Istituto di Clinica Neurologica, University of Trieste, Trieste, Italy.
Correspondence to Marino Zorzon, MD, Istituto di Clinica Neurologica dell'Università degli Studi di Trieste, Ospedale di Cattinara, Strada di Fiume, 447-34149, Trieste, Italy.
| Abstract |
|---|
|
|
|---|
Methods In a case-control study, we compared the prevalence of vascular risk factors in 64 TGA patients with 64 first-ever transient ischemic attack (TIA) control subjects and 108 normal community-based control subjects matched for age and sex. We prospectively studied the vascular events and mortality rates of the TGA cases and of the TIA control subjects. Then we compared the outcome of the two groups using actuarial analysis based on survival curves.
Results We did not find evidence of an increased risk of TGA associated with any vascular risk factor. In contrast to TIA control subjects, no TGA patient suffered stroke, myocardial infarction, or TIA during the follow-up period. Migraine was more common in TGA patients than in both normal and TIA control subjects. In three patients (4.5%), the TGA was eventually considered to be of epileptic origin.
Conclusions The results of our case-control and longitudinal studies point to the conclusion that TGA and TIA do not share the same etiology. Since half of our patients had a precipitating event in their history, it is reasonable to hypothesize that spreading depression may play a role in TGA. The significant positive association between migraine and TGA may support this hypothesis. Epilepsy may mimic TGA in a minority of cases.
Key Words: amnesia cerebral ischemia, transient prognosis risk factors
| Introduction |
|---|
|
|
|---|
The symptoms of TGA are usually typical and rather stereotyped, consisting of a sudden onset of transient inability to retain new information and to recall previous events for a variable period of time, generally occurring in middle-aged or elderly patients formerly in good health and without significant cardiac or cerebrovascular disease.4 5 6 7 These episodes usually last a few hours and then disappear completely, leaving a total or (rarely) a partial amnesia for the events that occurred during the attack and sometimes a short period of permanent retrograde amnesia.7 Because patients are aware of their memory impairment, they are often restless and worried, asking the same questions over and over again. Immediate memory, personal identity, and consciousness are well preserved, and the patients retain the ability to perform even complex routine tasks requiring visual and spatial skills. Other important systemic and focal neurological or neuropsychological deficits are absent, as are epileptic features.6 7 8
A great number of TGA attacks follow a series of potential precipitating factors and activities such as strenuous physical exercise, sexual intercourse, hot and cold baths or showers, emotional stress, driving or riding a motor vehicle, and medical procedures (especially angiography), but in most episodes, previous events are not noted.1 2 5 7 9 10 11 12 13 14 15
Although TGA has been described in association with several pathological conditions,16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 the most tenable hypotheses that have been put forward so far to explain the pathogenesis of TGA are transient ischemia in the inferomedial parts of the temporal lobes,11 12 31 38 44 45 46 epileptic discharge in the hippocampus,15 16 18 19 or recently, spreading depression of cortical electrical activity47 like that described in experimental animals.48
Most authors consider TGA a benign condition with a low risk of recurrence and of subsequent vascular events and mortality,1 2 5 7 11 49 50 51 52 53 54 but others report a high incidence of stroke or mental deterioration.31 44 46
These discrepancies are probably due to the lack of generally accepted diagnostic criteria, since some authors have included patients with unwitnessed episodes or with focal neurological signs during the attack. The adoption of strict diagnostic criteria for TGA7 55 in the recent literature demonstrated that the prognosis for pure TGA is excellent.2 49 50 51 53 54
The analytic epidemiology can help us ascertain the etiology of TGA by comparing the prevalence of vascular risk factors and the outcome of TGA patients with TIA patients.
This issue was addressed recently in a few case-control studies,2 45 49 53 56 but the question is still open; some authors lean toward a vascular etiology,45 53 while others reject it2 or consider it unlikely.49 56
To explain the etiology and prognosis of TGA, we studied the cerebrovascular risk factors and associated conditions and evaluated the outcome in a group of patients with TGA. Then we compared the prevalence of cerebrovascular risk factors in TGA case patients with TIA control and normal control subjects and the outcome in the TGA and TIA groups.
| Subjects and Methods |
|---|
|
|
|---|
Detailed information about the distinctive features of the episode were collected from the patient, his or her relatives, and the general practitioner: duration, potential precipitating factors and activities preceding the attack, accompanying neurological symptoms and signs, presence of cerebrovascular risk factors including previous TIA or stroke, associated neuropsychiatric disorders, and cardiovascular disease and other morbid conditions potentially correlated with TGA pathogenesis.
Patients who met the following criteria were considered for the study: (1) abrupt onset of inability to retain new information and retrograde amnesia of variable extension; (2) preservation of immediate memory and personal identity without cognitive impairment other than amnesia; (3) evidence provided by a reliable observer who witnessed the episode from the beginning; (4) absence of important accompanying focal neurological signs (only slight asymmetry of the reflexes was allowed) or epileptic features; (5) regression of symptoms within 24 hours except persistent amnesia for the period of the attack; and (6) no epilepsy, psychiatric illness, recent head trauma (in the 72 hours preceding the episode of TGA), progressive mental deterioration, or alcohol or drug abuse in past medical history.
These criteria, which coincide with those suggested by Caplan,7 were fulfilled by 67 patients. Thirteen patients were excluded: four had active epilepsy, four had unwitnessed episodes, three suffered from persistent amnesia, and two had a recent head injury. In three patients, the TGA was eventually considered to be of epileptic origin; consequently, they were excluded from the analysis. Therefore, the study population comprised 64 patients.
All patients were submitted by two of us (M.Z. and L.A.) to a standardized clinical investigation including full neurological and neurovascular examinations during the attack (24 patients) or within 24 hours after it (40 patients), since all the patients were hospitalized for a recent TGA episode.
The following laboratory examinations were performed whenever possible: routine blood tests, ECG, cardiological examination, chest x-ray, EEG, Doppler ultrasound study of the extracranial arteries, and CT scan of the head.
Afterward, the patients included in the study were prospectively reviewed annually by the same investigators who performed the initial examination. The follow-up was carried out with a structured interview planned to provide information about recurrence of TGA, vascular or nonvascular death, TIA, stroke, MI, or other vascular events and neurological diseases potentially associated with TGA such as epilepsy or migraine.
The questionnaire was completed during a medical examination or, when this was impossible, through a telephone call to the patient, relatives, or doctor. Patients reporting significant medical problems were visited, and their doctors were contacted for further medical details. No patient was lost to follow-up. According to age (±2 years) and sex, we matched each TGA patient with one patient affected by first-ever TIA who was referred to the Department of Neurology of the University of Trieste during the same period of presentation. TIA cases were selected from a personal series of one of the authors (M.Z.) who was maintaining regular follow-up. The chronologically nearest patient without any previous cerebral ischemic event was chosen. No other selection criteria were used. For each TGA patient, two age- (±2 years) and sex-matched normal community-based control subjects were selected from the registers of two general practitioners working for the National Health Service. The alphabetically closest names were taken. The 128 subjects were invited to join the study, and all willing participants were interviewed and examined by Dr Vitrani to obtain the same data as for the TGA cases. The response rate was 86% (110 persons), but two were discarded for insufficient data. No alternative second choice was selected. Hence, the final community-based control group consisted of 108 subjects, 47 men and 61 women with a mean±SD age of 61.8±7.6 years.
The following vascular risk factors and associated conditions were
considered in the case-control study: arterial hypertension
(currently being treated or found at least twice to have a systolic
blood pressure
160 and/or a diastolic blood pressure
>95), diabetes mellitus (currently being treated or fasting blood
sugar >7.77 mmol/L), smoking (>10 cigarettes daily),
hypercholesterolemia (fasting serum
cholesterol >5.69 mmol/L), ischemic heart disease
(angina or proven MI), peripheral vascular disease, atrial
fibrillation (ECG proven), cerebrovascular disease (past stroke or
TIA), and migraine (with or without aura).57
To estimate the relative risk for the potential risk factors, the odds ratio with 95% confidence intervals was calculated.58 If the 95% confidence interval included 1.0, the difference between the groups was considered to be nonsignificant at the 5% level.59 The outcome was assessed with actuarial methods based on survival tables, and their significance was tested according to the statistical package SPSS/PC+.
| Results |
|---|
|
|
|---|
TGA attacks lasted from 20 minutes to 20 hours. The mean duration of amnesia was 5.7±1.8 hours. Only 2 patients (3.1%) reported a previous episode of TGA. Potential precipitating factors preceding the attack were reported in half of the cases: physical exercise in 18 (28.1%), sexual intercourse in 7 (10.9%), emotional stress in 4 (6.3%), and hot bath in 3 (4.7%).
During the attack, 5 patients (7.8%) suffered from headache, 7 (10.9%) had high blood pressure, and 1 (1.6%) reported paresthesia in the left arm.
When examined during or shortly after the episode, 4 (6.3%) patients showed only slight asymmetry of the reflexes. No other significant neurological signs were found.
Interictal EEG was considered normal in 45 patients (70.3%) and abnormal in 19 (29.7%). Mild to moderate nonspecific abnormalities (slow activity) were present in 18 patients (3 bilateral, 3 right-sided, and 12 left-sided), and only 1 patient showed focal left temporal slowing without epileptic discharges.
EEG recordings were obtained during the attack from 5 patients: two were normal, and three showed right temporal focal theta activity, which disappeared in the EEG performed 2 days later.
Fifty patients (78.1%) underwent a CT scan of the head. CT was normal in 39 patients (60.9%), 4 (6.3%) showed mild diffuse cortical atrophy, 4 (6.3%) had slight ventricular enlargement, 1 (1.6%) showed mild cerebellar atrophy, and 2 (3.1%) had a hypodensity, left occipital in 1 and right deep temporal in the other, consistent with cerebral infarction.
Doppler ultrasound examination, performed in 28 (43.8%) patients, was normal in 27 and revealed flow turbulence in the left carotid bifurcation in 1 patient. Cerebral angiography was performed in 3 patients. It was normal in 1 patient, but in 2 others it showed a tight left carotid stenosis and an atherosclerotic plaque in the right carotid bifurcation, respectively.
The prevalence of principal cerebrovascular risk factors and associated
conditions in TGA cases was compared with that found in TIA and normal
control subjects (Table 1
).
|
Ischemic heart disease, smoking, and peripheral vascular disease were associated with a significantly reduced risk for TGA compared with TIA. The only factor significantly associated with an increased risk for TGA compared with TIA was migraine. Regarding hypertension, atrial fibrillation, diabetes mellitus, and hypercholesterolemia, there was no risk difference.
When comparing TGA patients with normal control subjects, again the only factor significantly associated with an increased risk for TGA was migraine, whereas smoking was associated with a significantly reduced risk for TGA. Regarding ischemic heart disease, hypertension, atrial fibrillation, diabetes mellitus, hypercholesterolemia, peripheral vascular disease, and cerebrovascular disease prevalence, there were no differences between the groups.
The mean period of follow-up in TGA patients was 45.6±35.7 months, with a total length of follow-up of 243 patient-years. The corresponding figures for TIA control subjects were 46.1±34.9 months and 245 patient-years.
During the follow-up period, three deaths occurred in the TGA group and
six deaths in the TIA group. None of the deaths of the TGA patients was
due to vascular causes. In the TIA control subjects, three deaths were
due to vascular causes and three to nonvascular causes. Vascular event
incidence and mortality rates in the TGA and TIA groups are illustrated
in Table 2
. The mortality rate in the TIA control
subjects was double that in the TGA patients. The actuarial
analysis of survival failed to show a significant difference
between the groups.
|
In the TIA control subjects, 5 fatal or nonfatal strokes, 5 TIAs, and 3
nonfatal MIs occurred during the follow-up. No vascular event was noted
in the TGA patients. Comparison of survival free from stroke, from TIA
and stroke, and from all vascular events (TIA, fatal or nonfatal
stroke, fatal or nonfatal MI) showed a significant difference between
the groups (P=.0469, P=.0066, and
P=.0041, respectively) (Fig 1
). The incidence
of major vascular events (stroke, MI, or vascular death) in the TIA
control subjects was significantly higher (P=.0317) than in
the TGA group (Fig 2
).
|
|
Recurrence of TGA was observed in 6 patients (9.4%) with a low average annual recurrence rate (2.5%). None of the TIA control subjects experienced a TGA episode during the follow-up.
A few months after presentation, 3 patients (4.5%) repeatedly experienced brief spells of amnesia that quickly subsided after antiepileptic treatment. CT scans and EEG recordings performed after the recurrences remained normal. Generalized convulsion or complex partial epilepsy were not reported. Epilepsy was not noted in TIA control subjects.
| Discussion |
|---|
|
|
|---|
There was a slight preponderance of women (sex ratio, 1.3), as already observed by others,3 10 45 46 54 but this result probably reflects the overrepresentation of women among the elderly in our region.
The mean duration of TGA attacks was about 6 hours, as confirmed by the data reported in two recent extensive studies.1 51 Very short spells and attacks longer than 12 hours were unusual.
In most series, the proportion of patients presenting with more than one episode of TGA is between 10% and 30%7 21 45 51 ; in our series, 97% of cases have been included after a single attack, and the observed frequency of TGA recurrence of 9.4% matches well with the earlier studies.
A potential precipitating factor has been identified in 50% of the cases. This figure is quite high if compared with that observed in retrospective studies,1 11 46 but it is comparable to or even lower than that reported among prospective collected cases.2 49 53
More than 30 years after Fisher and Adams4 5 used the term for the first time, the etiology and pathogenesis of TGA is still disputed. Some authors have suggested an epileptic origin on the basis of abruptness of onset, brevity of the attack, and reversibility of the symptomatology, combined with EEG abnormalities in some instances.5 15 16 18 19 However, the length of the attack (which usually lasts hours not minutes), the absence of impairment of consciousness and of other cortical functions other than amnesia, the lack of generalized convulsion, and repetitiveness do not render the epileptic origin very probable in the majority of cases. Furthermore, EEG recordings obtained during typical TGA spells are usually normal and very rarely show epileptiform discharges.7 60 61 62 63
Recently, Hodges and Warlow2 pointed out that in their series 7% of patients eventually developed epilepsy, and they concluded that epilepsy may initially mimic TGA in a not-negligible minority of patients.
In our series, epilepsy should be considered as an uncommon cause of TGA because during the follow-up, only three patients (4.5%) had brief repetitive spells of TGA, which soon disappeared after antiepileptic medication, making the diagnosis of epilepsy highly probable. The proportion of patients (4.5%) who were initially misdiagnosed as having a pure TGA in our study is in accordance with that of Hodges and Warlow,2 while in other prospective studies no patient had seizures during the follow-up.49 53
Most authors have accepted the hypothesis that TGA is cerebrovascular in origin. The favorite explanation has been a thromboembolic occlusion,10 11 12 13 31 38 44 45 46 but vasospastic, vasomotor, or hemodynamic mechanisms8 51 53 64 and an undescribed type of lacune6 have been proposed also.
Transient ischemia in the territory of the posterior cerebral artery,12 31 44 45 or alternatively of the anterior choroidal artery,10 causing dysfunction in the deep limbic system involved in memory, has been suspected of being responsible for TGA.
The thromboembolic cerebrovascular disease hypothesis is supported by a very high prevalence of vascular risk factors and by a high incidence of subsequent stroke and permanent memory defect in TGA patients observed in some studies.11 31 44 45 46 50 65 66
This view has been recently questioned; a well-conducted case-control study comparing TGA patients and both normal and TIA control subjects failed to demonstrate an increased risk of TGA associated with any of the cerebrovascular risk factors and showed very significant differences in the incidence of vascular events and mortality between TGA and TIA groups.2 This was also the case in the present study and in the Portuguese study49 ; Guidotti et al53 found a similar prevalence of vascular risk factors in TGA and TIA, but only TIA patients suffered stroke or MI during follow-up.
The results of the case-control studies published so far are contradictory with regard to etiology: some favor a vascular etiology, either through a thromboembolic45 or a vasospastic53 mechanism, whereas others reject it2 or consider it unlikely.49 56
By comparing 64 patients affected by TGA with 64 sex- and age-matched control subjects suffering from first-ever TIA, we found a significant difference in the prevalence of some vascular risk factors (ischemic heart disease, smoking, and peripheral vascular disease), indicating a negative risk for TGA patients associated with these factors.
In a comparison of TGA patients with 108 sex- and age-matched community-based normal control subjects, the analysis of relative risk showed a reduced risk for TGA patients associated with smoking and no excess risk associated with any of the other vascular risk factors.
The follow-up comparison of TGA and TIA groups in our study showed a significant difference in the incidence of stroke (fatal and nonfatal), TIA and stroke, and all vascular events (actually, none of the TGA patients had any vascular event during the follow-up), hence confirming the better prognosis of the TGA patients for subsequent vascular events.
The mortality (all causes, vascular and nonvascular) did not significantly differ in TGA and TIA groups.
A recurrence of the TGA episode occurred in six patients in the TGA group, but no patient in the TIA group experienced a TGA episode, thus confirming once more the remarkable difference in the natural history of the TGA and the TIA patients. The occurrences during follow-up of stroke, MI, and TIA only in TIA control subjects (while TGA patients did not suffer any of these) and the fact that TIA control subjects did not experience any TGA (while TGA patients did have recurrent TGAs during follow-up) heavily support the idea that TGA and TIA do not share a common etiology, which is in accordance with earlier well-conducted case-control studies.2 49
A cerebral infarct on CT scan was discovered in two (3.1%) of our TGA patients, but in one patient it was located in the occipital lobe, outside the structures involved in memory, and only in the other was it properly situated in the temporal lobe. Yet the temporal and causal relationship with the TGA episode remained unproven. Finally, Doppler ultrasound examination in our TGA patients was abnormal only in one (1.6%); this was recorded also by other authors who very rarely found abnormal Doppler ultrasounds of the extracranial arteries in TGA patients, in contrast to the findings in TIA patients.49 53
Many authors emphasized the association between TGA and migraine.2 21 22 24 40 67 68 69 Cerebral blood flow abnormalities, similar to those found in migraine and different from those seen in patients with TIA, were found after TGA episodes.21 Single-photon emission computed tomography studies showed bilateral temporal lobe hypoperfusion during TGA attack, but there is uncertainty whether the perfusion deficit is primary or secondary to a fall in cerebral metabolism.70 71 72 73 74
A reduction in cerebral blood flow, cerebral metabolic rate of oxygen, and tissue oxygen extraction in the temporal lobes during a TGA spell were documented in positron emission tomography studies, suggesting decreased metabolism.75 76 77 78
It is interesting to note that the phenomenon of spreading depression of cortical activity described by Leao48 produces similar changes in experimental animals79 80 ; recently it has been suggested that TGA and migraine share this common pathogenetic mechanism.47
In our series of patients with TGA, in accordance with other case-control studies,2 49 56 the prevalence of migraine (with or without aura) was significantly higher than in both TIA and normal control subjects. Therefore, migraine appears to be associated with an increased risk for TGA compared with either control group.
Those with migraine represent roughly a quarter of our TGA patients, therefore migraine is unlikely to be the only explanation of TGA for most of them.
In our opinion, it is remarkable that potential precipitating factors and activities are present in a large number of cases (50%), as previously reported by others.2 49 53 Since a coincidence seems rather unlikely, it can be hypothesized that intense sensory input may trigger spreading depression in the hippocampus47 or a vasomotor response1 49 leading to reversible functional ablation of memory structures.
In conclusion, our case-control data show that the prevalence of vascular risk factors is similar in the TGA group and in the normal control subjects, whereas the prevalence of some vascular risk factors (ischemic heart disease, smoking, peripheral vascular disease) is significantly greater in the TIA group.
The prospective longitudinal study shows that the number of vascular events in the TIA patients is significantly greater than in the TGA patients. Although there are not significant differences in mortality, the overall low mortality rate during the follow-up period proves the benignity of TGA, in contrast to the natural history of TIA. The results confirm that it is questionable whether TGA and TIA share the same etiology.
Since half of our patients had a precipitating event in their history, it is reasonable to hypothesize that spreading depression may play a role in TGA.
Finally, we want to emphasize the significant positive association between migraine and TGA, which may have an etiologic implication at least in some cases, and that epilepsy can mimic TGA attacks in some cases.
| Selected Abbreviations and Acronyms |
|---|
|
| Acknowledgments |
|---|
Received February 9, 1995; revision received May 18, 1995; accepted May 19, 1995.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. M. Schott, M. D. Filippo, and P. Calabresi ISCHEMIC BILATERAL HIPPOCAMPAL DYSFUNCTION DURING TRANSIENT GLOBAL AMNESIA Neurology, May 13, 2008; 70(20): 1940 - 1941. [Full Text] [PDF] |
||||
![]() |
T. Bartsch, K. Alfke, R. Stingele, A. Rohr, S. Freitag-Wolf, O. Jansen, and G. Deuschl Selective affection of hippocampal CA-1 neurons in patients with transient global amnesia without long-term sequelae Brain, November 1, 2006; 129(11): 2874 - 2884. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Menendez Gonzalez and M. M. Rivera Transient global amnesia: increasing evidence of a venous etiology. Arch Neurol, September 1, 2006; 63(9): 1334 - 1336. [Full Text] [PDF] |
||||
![]() |
E. S. Roach Transient global amnesia: look at mechanisms not causes. Arch Neurol, September 1, 2006; 63(9): 1338 - 1339. [Full Text] [PDF] |
||||
![]() |
P. Quinette, B. Guillery-Girard, J. Dayan, V. d. l. Sayette, S. Marquis, F. Viader, B. Desgranges, and F. Eustache What does transient global amnesia really mean? Review of the literature and thorough study of 142 cases Brain, July 1, 2006; 129(7): 1640 - 1658. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. -P. Chung, H. -Y. Hsu, A. -C. Chao, F. -C. Chang, W. -Y. Sheng, and H. -H. Hu Detection of intracranial venous reflux in patients of transient global amnesia Neurology, June 27, 2006; 66(12): 1873 - 1877. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. J. Bortolon, M. R. Weglinski, and J. Sprung Transient Global Amnesia After General Anesthesia Anesth. Analg., September 1, 2005; 101(3): 916 - 919. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Nakada, I. L. Kwee, Y. Fujii, and R. T. Knight High-field, T2 reversed MRI of the hippocampus in transient global amnesia Neurology, April 12, 2005; 64(7): 1170 - 1174. [Abstract] [Full Text] [PDF] |
||||
![]() |
K Winbeck, T Etgen, H G von Einsiedel, M Rottinger, and D Sander DWI in transient global amnesia and TIA: proposal for an ischaemic origin of TGA J. Neurol. Neurosurg. Psychiatry, March 1, 2005; 76(3): 438 - 441. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Hirabayashi, M. Shimizu, S. Kohara, and Y. Shinohara Activated platelets in transient global amnesia and TIA Neurology, August 24, 2004; 63(4): 736 - 738. [Abstract] [Full Text] [PDF] |
||||
![]() |
O. Sedlaczek, J. G. Hirsch, E. Grips, C. N.A. Peters, A. Gass, J. Wohrle, and M. Hennerici Detection of delayed focal MR changes in the lateral hippocampus in transient global amnesia Neurology, June 22, 2004; 62(12): 2165 - 2170. [Abstract] [Full Text] [PDF] |
||||
![]() |
S.-T. Chen, L.-M. Tang, T.-H. Lee, L.-S. Ro, R.-K. Lyu, and S.-T. Chen Transient Global Amnesia and Amaurosis Fugax in a Patient with Common Carotid Artery Occlusion: A Case Report Angiology, March 1, 2000; 51(3): 257 - 261. [Abstract] [PDF] |
||||
![]() |
H. Nakane, S. Ibayashi, K. Fujii, S. Sadoshima, K. Irie, T. Kitazono, and M. Fujishima Cerebral blood flow and metabolism in patients with silent brain infarction: occult misery perfusion in the cerebral cortex J. Neurol. Neurosurg. Psychiatry, September 1, 1998; 65(3): 317 - 321. [Abstract] [Full Text] |
||||
![]() |
A. R. Woolfenden, M. W. O'Brien, R. E. Schwartzberg, A. M. Norbash, and D. C. Tong Diffusion-Weighted MRI in Transient Global Amnesia Precipitated by Cerebral Angiography Stroke, November 1, 1997; 28(11): 2311 - 2314. [Abstract] [Full Text] |
||||
| ||||||||||