Jugular Veins in Transient Global Amnesia
Background and Purpose—Transient global amnesia (TGA) has been associated with an increased prevalence of internal jugular valve insufficiency and many patients report Valsalva-associated maneuvers before TGA onset. These findings have led to the assumption of hemodynamic alterations in intracranial veins inducing focal hippocampal ischemia. We investigated this hypothesis in patients with TGA and control subjects.
Methods—Seventy-five patients with TGA and 75 age- and sex-matched healthy subjects were enrolled into a cross-sectional study. Extracranial and transcranial high-resolution venous echo-color-Doppler sonography was performed blindly in all patients and control subjects. Blood flow direction and velocities were recorded at the internal jugular veins, basal veins of Rosenthal, and vein of Galen, both at rest and during Valsalva-associated maneuvers.
Results—Mean age of patients with TGA was 60.3±8.0 years (median, 60 years; range, 44–78 years); 44 (59%) were female (female/male ratio: 1.42). Internal jugular valve insufficiency (left, right, or bilateral) was found to be more frequent in patients with TGA than in control subjects: 53 (70.7%) versus 22 (29.3%; P<0.05). Blood flow velocities in the deep cerebral veins of patients with TGA did not differ from control subjects both at rest and during Valsalva-associated maneuvers. Intracranial venous reflux was neither observed in patients with TGA nor in control subjects despite unilateral or bilateral internal jugular valve insufficiency during prolonged and maximal Valsalva-associated maneuvers.
Conclusions—This study, although confirming the association between TGA and internal jugular valve insufficiency, challenges the hypothesis that cerebral venous congestion plays a significant role in the pathogenesis of TGA.
See related article, p 2275.
Despite its benign course, transient global amnesia (TGA) remains one of the most intriguing syndromes in clinical neurology for its striking onset and obscure pathogenesis.1 It presents classically with an abrupt and severe anterograde amnesia and a milder retrograde amnesia, accompanied by repetitive questioning, that worries patients and their families.2,3 Patients do not have focal neurological symptoms or cognition impairment, but they are disoriented to time and place. TGA usually affects patients between the ages of 40 and 80 years and lasts <24 hours, most often between 1 and 8 hours. Most of patients experience a single episode with a low recurrence rate.4 Diagnosis can be made safely in the presence of a characteristic history and by applying the criteria proposed by Hodges and Warlow in 1990.5
Given the memory impairment, an involvement of temporal lobe structures including the hippocampus has long been suggested.5 Recent MRI data in patients with TGA suggest that focal diffusion-weighted imaging of hyperintense lesions in the CA1 field of the hippocampal cornu ammonis might represent the pathological correlate of a transient perturbation of hippocampal function.6 These acute hippocampal cytotoxic-like lesions have been detected in up to two thirds of patients with TGA.7 Nonetheless, the nature of these MRI alterations is still unknown. In fact, various factors have been suggested to be involved in the pathophysiology of TGA such as focal ischemia (especially in the posterior circulation), migraine attacks, seizure activity, and venous flow abnormalities.8,9 In particular, valve incompetence of the internal jugular vein (IJVI) has been documented in the majority of patients with TGA and it has been posited by many studies and reports that TGA may result from cerebral venous congestion due to a venous reflux during Valsalva-like activities before symptom onset.10–15 However, this hypothesis does not match with some clinical TGA characteristics such as age of onset and recurrence rate. Besides, to prove a direct link between IJVI and TGA, we would expect to observe an alteration of cerebral blood flow in the veins draining the hippocampus, namely the basal veins of Rosenthal and the vein of Galen, at least during Valsalva-associated maneuvers (VM). These vessels, which are highly suspected to be involved in TGA, can be reliably studied by transcranial ultrasound. Surprisingly, this issue has never been addressed so far. For this very reason, the aim of our ultrasound study was to assess, in a large sample of patients with TGA and in age- and sex-matched normal control subjects, the prevalence of IJVI and its impact on cerebral venous hemodynamics at rest and during VM.
Seventy-five patients with their first-ever episode of TGA seen at the Department of Neurological Sciences of the University of Padua and at the Division of Neurology of the Regional Hospital of Treviso were enrolled into this study. The diagnosis was achieved in agreement with the Hodges and Warlow diagnostic criteria.5 The diagnostic workup included: (1) a neurological examination; (2) biochemical blood tests; (3) brain CT or MRI; (4) electroencephalography; and (5) a transcranial and extracranial high-resolution color-coded Doppler sonography. Fifteen patients were assessed by ultrasound at the time of being still symptomatic. Seventy-five subjects matching for age and sex with the patients with TGA represented the control group (NC). All control subjects underwent extracranial high-resolution color-coded Doppler sonography.
Venous Ultrasound Examination
Patients and control subjects underwent a complete extracranial and intracranial venous ultrasound assessment at the Neurosonology Laboratory of our department. All examinations were performed by a high-resolution color-coded duplex sonography scanner (PhilipsiU22) using a high-frequency (5–10 MHz) linear probe for the cervical veins and a low-frequency (1–3 MHz) phased-array probe for the intracranial veins. This instrumentation and our personal setting allowed to achieve excellent B-mode imaging, especially of the brain parenchyma; this represents the basis for a correct analysis of the data collected in the color and Doppler modes. The examination was performed by an experienced neurosonographer blinded to the patient's diagnosis, in the same room, in a quiet atmosphere, with the subjects lying in a supine position. Blinding was further assured by avoiding any communication between the patient/control subject and the neurosonographer.
For extracranial examination,16 the patient was asked to place his or her head in a straight position to avoid flow alterations caused by unilateral or bilateral venous outflow obstruction. A large amount of gel was placed on the skin of the patient and great care was taken not to compress the cervical veins when the probe was applied over the neck to obtain reliable velocity measurements. Before recording the hemodynamic data, the patient was asked to rest in that position for at least 2 minutes and take several deep breaths to allow blood redistribution in the venous system. The internal jugular veins were examined by using the transverse and the longitudinal cervical insonation planes; the direction of flow was analyzed by comparing the color of the flow and the direction of the pulsed wave in the internal jugular vein with that of the satellite carotid artery. Longitudinal B-mode insonation of the internal jugular vein in its caudal segment served to visualize the inferior jugular bulb and the jugular valve and look for malformations in the internal jugular veins (septum, valve malformation) that might reduce or block venous outflow, even during inspiration. The lack of a Doppler signal in a vein despite several deep inspirations defined an outflow occlusion. IJVI was assessed on both sides after repeated training of the patients and subjects during 3 maximal VMs (duration >5 seconds). An adequate VM was confirmed by a clearly visible increase of internal jugular vein diameter (approximately ≥1 cm) in the B-mode image. In case of valve incompetence, flow reversal was documented during the Valsalva maneuver. Reflux was defined as a reverse flow assessed in the respiratory pause for a duration >0.88 seconds.10
For transcranial examination,17 the system settings were adjusted for the analysis of low-velocity signals; therefore, the filters were switched off and the pulse repetition frequency was reduced for better venous vessel detection. The patient was examined in the supine position through the transtemporal bone window. We assessed the paired basal veins of Rosenthal (BV) and the unpaired vein of Galen. Hemodynamic data were collected from these veins only after having identified them correctly. The BV was insonated in its middle segment using the axial midbrain plane, whereas more distal parts were seen in the thalamic plane. In its proximal segment, the BV Doppler signal is usually found lateral of the P2 segment of the posterior cerebral artery with a flow direction toward the probe, whereas in its distal segments, it lies medial and superior to the P2 posterior cerebral artery and P3 posterior cerebral artery segments with a flow direction away from the transducer. The unpaired vein of Galen was visualized in an axial thalamic plane posterior to the hyperechogenic pineal gland; its flow direction is away from the probe. By following the signal of the BV until its junction with the unpaired vein of Galen, the contralateral BV is found; its flow direction is toward the transducer.
Blood flow direction and velocity of all these veins were recorded both at rest and during VM, taking great care not to measure at junctions with other vessels, because at these points, venous flow velocities could vary greatly. With regard to VM, we recorded data during the whole VM and reported velocity values during the last 3 seconds before exhalation, when maximal venous congestion is expected. A reflux was defined as a reverse flow of any duration at any time.
Clinical and demographic data are expressed as the mean±SD. Differences between the 2 groups were tested for significance with the one-way analysis of variance and, when necessary, with the 2-sided Fisher exact test. Significance was inferred at P<0.05.
Standard Protocol Approval and Patient Consent
The study was approved by our local ethics committee and informed consent was obtained for all patients and control subjects.
Demographic and Clinical Data
The mean age of the 75 patients with TGA was 60.3±8.0 years (median, 60 years; range, 44–78 years); 44 (59%) were female and 31 (41%) were male (female:male ratio: 1.42). The NC was composed of 44 females and 31 males (female/male ratio: 1.42) with a mean age of 60.2±7.9 (median, 60 years; range, 44–78 years). With regard to vascular risk factors, there was no difference between the 2 groups in terms of the rates of arterial hypertension, smoking, diabetes mellitus, hypercholesterolemia, atrial fibrillation, ischemic heart disease, cerebrovascular disease, carotid disease, intracranial atherosclerosis, and peripheral artery disease. Patients with TGA were more likely to have migraine (with or without aura; 29% versus 13%; P<0.0001) compared with NC. Trigger events were present in 54% of our patients with TGA; sexual intercourse and emotional stress were most prevalent (Table 1).
IJVI (left, right, or bilateral) was found in a significantly higher number of patients with TGA compared with NC: 53 (70.7%) versus 22 (29.3%).
BV and unpaired vein of Galen were detected in 100% and 96% (72 of 75) of patients with TGA versus 100% and 94.7% (71 of 75) of NC.
Velocity values at rest were within the normal range in both study groups and in those patients with IJVI. In patients with/without IJVI versus normal control subjects, there was no difference of venous blood flow velocities both at rest and during VM. Venous reflux was not found either in patients or in control subjects (Table 2).
The main finding of this study is the absence of any intracranial venous hemodynamic alteration in patients with TGA (ie, no reflux and no differences of venous blood flow velocities compared with control subjects) both at rest and during VM. This result was also observed in patients with TGA with IJVI, implying that intracranial veins do not seem to be directly involved in the pathogenesis of this syndrome. Several reports in which patients with TGA underwent ultrasound assessment suggested that extracranial venous abnormalities such as IJVI might play a key role in determining a cerebral venous engorgement leading to temporal–mesial dysfunction.10–15,18,19 Actually, none of these studies directly investigated cerebral venous hemodynamics, thus leaving a very important gap between their findings and the pathogenic hypothesis.
Indeed, this simple hypothesis does not match with some clinical observations; TGA appears only once during the life of most of these patients and VM does not always precede clinical onset. In more than half of our patients we identified a possible trigger factor confirming previous literature; nonetheless, their extra- and intracranial ultrasound findings did not differ from normal control subjects or from patients with TGA without trigger factors. Furthermore, most of the trigger factors represent ordinary activities so that their detection should not be a surprise when carefully interviewing patients with TGA. Consequently, it is difficult to explain why such usual activities are not able to continuously replicate TGA symptoms during a patient's life. None of our patients presented a relapse nor were TGA symptoms triggered by a sustained VM performed during our ultrasound investigation. Similarly, no data in the literature showed any association among heart failure, acute or chronic pulmonary hypertension, and TGA.20 Therefore, the role of trigger factors in TGA pathogenesis appears questionable.
Among vascular risk factors, only migraine was more frequent in patients with TGA compared with control subjects, confirming previous data4,21,22 and supporting the alternative hypothesis of TGA being a result of a cortical spreading depression. However, a cortical spreading depression in the hippocampus has never been shown in human beings, waning the role of this pathogenic mechanism.1 Moreover, only a minority of patients with TGA are migraineurs (<30% in our study). It might be suggested that several factors concur to TGA onset; nonetheless, in this study, patients with TGA with migraine did not show any difference with regard to IJVI prevalence or intracranial hemodynamics.
Beyond any doubt, this study confirmed the increased prevalence of IJVI in patients with TGA, yet there is a difficulty to link within a coherent pathogenic mechanism the 2 mentioned facts: long-lasting IJVI and a “mostly single episode” disease. In this context, our finding of normal intracranial hemodynamics even during VM further shuffles the cards and challenges the likelihood of the hypothesis forwarded by Nedelmann et al.10 In fact, an increased prevalence of IJVI does not imply its direct correlation with the onset of TGA. According to Lewis' hypothesis,8 during VM, IJVI might exert a pressure wave traveling upward from the jugular veins, thus modifying the intracranial pressure and intraluminal venous pressure rather than change the flow velocity. Both intracranial pressure and intraluminal venous pressure are components of cerebral hemodynamics. We did not measure intracranial pressure because invasive intracranial pressure monitoring is not feasible in patients with TGA nor in healthy control subjects. However, in the Starling resistor model, intracranial pressure increase should lead to reduced venous outflow. A complete overlap of blood flow velocities measured during VM in patients and control subjects with and without IJVI testifies that a retrograde venous congestion had no significant effect.
A possible criticism to our study is that a transient change in cerebral venous flow direction and/or velocity, secondary to IJVI, might precede TGA onset, and consequently an ultrasound examination performed even in the acute stage of TGA might miss it. However, this hypothesis does not explain why these venous alterations have always disappeared at ultrasound examination, because we did not observe them either during or soon after an TGA episode despite the persistence of IJVI.
Recently, an ultrasound study on arterial and venous intracranial flow during VM suggested the presence of an active regulation of cerebral venous tone,23 which acts independently from cervical venous hemodynamics. These new elements along with our findings shift the attention from extracranial to intracranial venous assessment opening a possible new scenario in the pathophysiology of TGA. The aim of future studies should focus on the assessment of intracranial venous hemodynamics during the acute stage of TGA to discriminate innocent bystanders from the true cause.
- Received February 14, 2012.
- Revision received May 28, 2012.
- Accepted May 30, 2012.
- © 2012 American Heart Association, Inc.
- Quinette P,
- Guillery-Girard B,
- Dayan J,
- de la Sayette V,
- Marquis S,
- Viader F,
- et al
- Hodges JR,
- Warlow CP
- Bartsch T,
- Alfke K,
- Stingele R,
- Rohr A,
- Freitag-Wolf S,
- Jansen O,
- Deuschl G
- Schreiber SJ,
- Doepp F,
- Klingebiel R,
- Valdueza JM
- Chung CP,
- Hsu HY,
- Chao AC,
- Chang FC,
- Sheng WY,
- Hu HH
- Stolz E,
- Kaps M,
- Kern A,
- Dorndorf W
- Cejas C,
- Cisneros LF,
- Lagos R,
- Zuk C,
- Meriso SF
- Caplan L