Comprehensive Ultrasound Assessment of the Craniocervical Circulation in Transient Global Amnesia

Structural changes and metabolic stress have been reported on diffusion‐weighted magnetic resonance imaging in the cornu ammonis 1 area of the hippocampus in patients with transient global amnesia (TGA), but a consensus on pathogenesis is still lacking. The aim of our study was to perform a comprehensive ultrasound analysis of the cerebrovascular circulation in our population of patients with TGA.

2][3] The aim of this research was to perform an ultrasound analysis of the cerebrovascular circulation in patients with TGA to assess the importance of the vascular mechanism in the pathogenesis of TGA in our population.

Materials and Methods
In a prospective study, we included 100 adult patients (18 years) with TGA, who were treated at the Neurology Clinic, Clinical Center of Serbia, from January 1, 2008, to January 1, 2015.The inclusion criterion was a diagnosis of TGA based on the diagnostic criteria of Hodges and Warlow. 4The exclusion criteria were diagnosis of a transient ischemic attack (TIA), acute stroke, nonconvulsive epileptic status and epileptic amnesia, hypoglycemia, head trauma, and psychiatric illness according to Diagnostic and Statistical Manual of Mental Disorders, fourth edition, classification. 5The control group comprised 50 age-and sex-matched inpatients at the Neurology Clinic who were treated in the same period for diseases of the peripheral nervous system and were selected according to the same exclusion criteria.The study was approved by the Review Board of the Neurology Clinic, Clinical Center of Serbia, and Ethics Committee of the Clinical Center of Serbia.All participants gave informed consent for participation in the study.The study was conducted in accordance with the World Medical Association Declaration of Helsinki.
Data on the clinical presentation, emotional stress or physical exertion with the Valsalva maneuver before episode onset, duration of amnesia, blood pressure during the TGA episode, previous cerebrovascular disorders, previous TGA, and vascular risk factors were obtained by interviews with the patients and their companions.All patients underwent the same diagnostic protocol, including a neurologic examination and laboratory analysis (full blood count, C-reactive protein, biochemical analysis, lipid levels, coagulation screening status, thyroid status, and immunologic analysis if it was indicated to exclude vasculitis).Brain computed tomography (CT) on a 64-slice multidetector row CT scanner was done on the first assessment in all patients with TGA and followed by a control CT scans in 35% within the first week.Magnetic resonance imaging examinations on a 1.5-T scanner (Avanto; Siemens AG, Erlangen, Germany) with T1-weighted, T2-weighted, and fluidattenuated inversion recovery sequences was done in 85% of patients with TGA within a month.Computed tomography of the brain was performed in the control group primarily (60%) and MRI optionally (40%) within a month.Standard electroencephalography, an ocular fundi examination, and a cardiac examination with electrocardiography and transthoracic echocardiography were done in all participants.The TGA group also underwent electroencephalography after sleep deprivation.Participants with a suspected right-to-left cardiac shunt based on positive bubble test results on transcranial ultrasound imaging underwent transesophageal echocardiography (TEE).All ultrasound examinations were performed within 7 days of TGA onset.
All participants underwent the following ultrasound scans, performed by blinded trained sonologists: 1. Color triplex examination of cervical segments of the carotid and vertebral arteries 6 with Mannheim criteria for intima-media thickness and atheromatous plaques. 7The competence of the internal jugular vein (IJV) valve was investigated by assessment of venous flow at rest (regular breathing) and at deep inspiration with the Valsalva maneuver; the occurrence of venous reflux was treated as a positive test result 8 (a-10; Hitachi Aloka Co, Ltd, Tokyo, Japan; 7.5-14-MHz transducer).went an ultrasound examination of the veins in the extremities, pelvis, and abdomen 14 (Hitachi Aloka a-10; 7.5-14-MHz transducer).
Results were tabulated.Statistical analyses included a descriptive analysis and a logistic regression analysis.P < .05 was considered statistically significant.

Results
The study groups were comparable in terms of age, with mean ages of 62 years in the TGA group and 61 years in the control group (P 5 .569;Table 1).Female participants were more common in both groups (67% in the TGA group and 60% in the control group; P 5 .549).Among patients with TGA, 41% had a university degree education; 48% graduated from high school; and 11% had an elementary school level of education.
The clinical presentation of TGA comprised a typical clinical picture in all participants.The amnesia duration ranged from 2 to 15 hours (mean 6 SD, 5.9 6 3.4 hours).The neurologic examination revealed normal findings in all patients.In most of the patients with TGA (97%), increased blood pressure was registered at the beginning of the TGA episode, ranging from 150/90 to 220/120 mm Hg, with a mean value of 165/95 mm Hg.Multiple old lacunar infarcts were detected on MRI scans in 6 of 100 patients with TGA (and 2 of 50 controls), which were localized: frontoparietotemporal bilaterally in 3 patients with TGA, in the region of the basal ganglia in 2, and paraventricular bilaterally in 1.
Table 1 shows demographic characteristics, risk factors, and precipitating factors in both study groups.Most patients with TGA (89%) were previously treated for arterial hypertension, but in 8% of these patients, the high pressure was measured the first time (altogether 97%).The patients with arterial hypertension had an 8 times higher risk of TGA compared to those without arterial hypertension (odds ratio [OR], 8.09; P < .001).The Duration of the hypertension diagnosis ranged from 2 to 29 years (mean, 7.7 6 6.1 years).A longer duration of hypertension was more frequently registered in the TGA group than the control group (P < 0.01).Hypertensive changes on the ocular fundi were diagnosed in 76 of 100 patients with TGA compared to the control group (19 of 50), and they were significantly associated with an increased risk of TGA (OR, 2.31;  1).
With regard to precipitating factors, a previous Valsalva maneuver was strongly associated with a TGA episode (OR, 5.85; P 5 .001); it was the consequence of lifting heavy objects (19 of 39), strain at defecation (17 of 39), and cough with prolonged expectoration (3 of 39).Emotional stress was associated with a TGA episode at the level of a statistical trend (P 5 .099;Table 1).The emotional stress comprised a reaction to receiving unpleasant news (15 of 35), tumultuous interpersonal discussion (9 of 35), fear of imminent medical intervention (7 of 35), and sexual intercourse (4 of 35).
The laboratory analysis performed at admission to the hospital did not reveal any parameter associated with an increased risk of TGA (P > .050for all; Table 2).
There were no significant differences between TGA and control groups in common carotid artery intimamedia thickness values, frequency of carotid plaques, or severity of stenosis on the carotid artery examination (Table 3).The plaques in the carotid arteries were fibrocalcified and stable, except that the plaques were lipid and unstable in 3 of 100 patients with TGA and 1 of 50 controls.The mean stenosis rates in the carotid arteries were 28.7% 6 11.7% in the TGA group and 29.3% 6 10.9% in the control group, including 8 patients with diabetic polyneuropathy.The mean systolic velocity in both internal carotid arteries was 73 6 26 cm/s in the TGA group; it was 37 6 14 cm/s in both vertebral arteries; and they were within normal age-defined ranges.The logistic regression analysis showed that the presence of an incompetent IJV valve (54% in the TGA group) was significantly associated with an increased risk of TGA (OR, 4.16; P < .001;Table 3).No patients with TGA had a diagnosis of peripheral venous thrombosis during a TGA episode, based on clinical and ultrasound examinations.
The mean middle cerebral velocity was 45 6 10 cm/s, and the basilar artery velocity was 29 6 8 cm/s in the TGA group (in the range of physiologic values).No intracranial stenosis in blood vessels of the circle of Willis circle was identified in the TGA and control groups.The mean PI values were within the normal range in both the middle cerebral and basilar arteries, and no significant differences were detected between study groups (Table 3).The mean breath-holding index values were within the normal range in both groups, and no difference was detected between the groups (Table 3).A small number of patients in both groups had microembolic signals, with no statistically significant difference between the groups (Table 3).Also, no differences were registered with regard to the number of patients with positive bubble test results or pathologic TEE results between groups (Table 3).

Discussion
We report a group of patients with TGA who underwent detailed a neurosonologic examination comprising extracranial and cranial vessel assessments.In our study, patients with TGA were older, as reported in the literature: typically between 50 and 70 years. 2,15Women had TGA more frequently, which was reported by other authors as well. 16,17It was observed that our patients with had higher education mostly, which corresponds to a high level of previous well-informed treatment of vascular risk factors.There were rare references in the literature about the professions of patients with TGA, but there are data on an obsessive-meticulous personality structure, emotional hypersensitivity, and the importance of psychophysical stress for the occurrence of TGA. 15,17,18We found that emotional stress as a precipitating factor increased the risk of TGA, although at the level of a statistical trend only.It is likely that emotional stress is associated with an acute rise of blood pressure, precipitating a TGA episode.We found a significant proportion of patients with TGA in whom procedures with the Valsalva maneuver immediately preceded the onset of TGA.The recurrence of TGA was rare, and the literature cited data between 6% and 10%, which is in accordance with the data in our study (10%). 2 During the episode of TGA, 97% of patients had acutely increased blood pressure.The presence of other risk factors was significantly less frequent or controlled well by medications.In some studies, analogous results were encountered in relation to vascular risk factors. 18,19n the same context, there were studies that compared the vascular risk factors in TGA and TIA; the most common risk factors were arterial hypertension and hyperlipidemia, but they were found considerably less often in TGA in comparison with TIA. 20On the other hand, some authors listed very rare occurrences of vascular risk factors in patients with TGA, arguing even that the mechanism of ischemic TGA should be rejected. 17,21here is also evidence that compared with patients with migraine or TIA, those with TGA do not seem to face a heightened risk of stroke. 22he signs of atherosclerosis were not particularly prominent in the large arteries of the neck and brain in our patients with TGA.Approximately half of the patients with TGA did not have atheromatous plaques in the carotid arteries, whereas the other half had only mild to moderate carotid atherosclerosis.4][25] Also, in accordance with other reports, no significant difference between study groups was registered with regard to the velocities of the main neck and cerebral arteries of the circle of Willis, which were within the physiologic range. 24,25he rare occurrence of microembolic signals in the patients with TGA in our study indicated that embolism had no essential role in the development of TGA.Although the detection of microembolic signals in patients with TGA has rarely been the subject of research by other authors, one recent study showed no significant occurrence of microembolic signals in patients with TGA as well. 24In our study, the same was related to transcranial color Doppler detection of a rightto-left cardiac or pulmonary shunt, thus excluding the importance of paradoxical embolism in the pathogenesis of TGA, which could be precipitated by a frequently reported Valsalva maneuver.Some previous investigations have shown a significantly higher incidence of a patent foramen ovale in patients with TGA, such as 55% in TGA compared to 27% in controls in a study by Kl€ otzsch et al, 27 supporting the role of paradoxical embolism of terminal branches of the basilar artery in the development of TGA. 28,29Recent investigations indicated no significant difference in the incidence of a patent foramen ovale in patients with TGA in relation to controls, as in our study; a patent foramen ovale was actually significantly less frequent in patients with TGA than those with TIA. 20,30n this examination, normal breath-holding index values in most of the patients with TGA excluded impairment of vasomotor reactivity of cerebral vessels and confirmed patency of large and small brain arteries.
Maybe the results would have been different if the breath-holding index had been determined at the beginning of TGA episodes, while blood pressure was elevated.In the literature, cerebral small blood vessel disease has been associated with TGA, although with no clear estimate of its importance. 15,31][34] Numerous studies of venous ultrasound and MRI examinations have shown the significant frequency of IJV valve failure in patients with TGA: from 50% to 85%, both bilaterally or unilaterally. 8,16,35,36Our previously published pilot study also indicated that 55% of patients with TGA had IJV valve incompetence. 37The results of this study, with double the number of patients with TGA, confirmed that more than half (54%) had IJV valve incompetence, which was significantly more than the control group (P < .01).Retrograde venous flow through the IJV causes congestion of the venous sinuses, Rosenthal basilar vein, and veins in the mesial part of the temporal lobe and hippocampus, structures that are relevant for memory processes. 15,35,36Venous congestion might have led to the occurrence of small venous thrombosis in the mesial part of the temporal lobe, particularly the hippocampus, resulting in TGA. 16,38The problem with this interpretation is that the chronic venous incompetence of the IJV valve should lead to more frequent recurrences of TGA, but it was known that the TGA relapsed rarely. 39Recent research also did not find evidence of intracranial IJV reflux on time-of-flight MR angiography. 26,40n recent years, the use of diffusion-weighted MRI and MR spectroscopy revealed metabolic disturbances in vulnerable neurons of the hippocampus, particularly in the CA1 region of the cornu amonis. 2 These disorders were shown as small round hyperintensities (1-3 mm) on diffusion-weighted MRI, the largest occurring between 48 and 72 hours from the onset of TGA and then gradually withdrawing; after 30 days, these changes were no longer detectable. 2,41Diffusion-weighted MRI showed restricted diffusion in these changes, whereas MRI spectroscopy revealed an increased amount of sodium aspartate as a sign of anaerobic glycolysis.These findings indicated acute and short-term metabolic stress in the CA1 region. 2,3,42We did not find typical MRI lesions in the temporal lobe, just old lacunar lesions of different distributions.In this study, we did not have the purpose or the ability to follow the evolution of MRI changes.Our patients had MRI from 2 to 4 weeks after the start of TGA episodes.
Can the results of our study be linked with modern knowledge about TGA? Based on the results of our comprehensive clinical and ultrasound research, the strongest connection with TGA in our data set of patients was the presence of incompetent IJV valves (as a possible basis for venous stasis in the veins and venules of the brain) and the sudden increase of blood pressure during a TGA attack (as a cause of acute vasospasm of the cerebral arteries and arterioles).
In conclusion, our ultrasound examination did not detect significant structural changes in the main arteries of the neck, the large arteries of the brain, or the small arteries of the brain.An emboligenic mechanism was excluded.The patients with TGA did not have the typical risk factors for cerebrovascular disease, or they were controlled well by medication.The only exception was that the patients had a jump in blood pressure during TGA attacks in the setting of IJV valve incompetency.New research should clarify whether these simultaneous functional circulatory changes have relevance for metabolic stress in the cornu ammonis of the hippocampus, especially during the early hours of a TGA attack.

Table 1 .
Demographic Characteristics, Risk Factors, and Precipitating Factors in the TGA and Control Groups Data are presented as mean 6 SD and number (percent) where applicable.CI indicates confidence interval.

Table 2 .
Laboratory Parameters in the TGA and Control Groups