Transient global amnesia following transoesophageal echocardiography

Abstract

Transient global amnesia (TGA) is characterised by the sudden onset of isolated anterograde amnesia, which resolves within 24 hours. Here, we discuss the case of a 63-year-old woman who underwent a transoesophageal echocardiogram (TOE) as part of her workup for pulmonary hypertension. She was well on the morning of the procedure, and following consent, underwent transoesophageal echocardiography without sedation. The procedure was uncomplicated with normal observations throughout, confirming a suspected secundum atrial septal defect. Immediately following oesophageal extubation, it was noted that the patient was disoriented. The physical neurological examination was unremarkable. Urgent MRI of the brain showed normal anatomy; a diagnosis of TGA was made. Within 10 hours of onset, the patient was back to her baseline. Isolated anterograde amnesia following transoesophageal echocardiography should raise the clinical suspicion of TGA. Prompt clinical examination and support from other specialties are paramount in making the right diagnosis.

Background

Transoesophageal echocardiography (TOE) is a relatively safe and well-tolerated imaging technique that provides high-resolution live imaging of the heart. Numerous studies are done on an outpatient basis, and patients have the option to undergo the procedure with or without sedation. Here, we describe an episode of transient global amnesia (TGA) triggered by TOE.

Case presentation

A 63-year-old woman was seen in respiratory clinic following three discrete episodes of bronchopneumonia within the space of 16 months. Of note, she had no significant medical history. She was a non-smoker, and had no relevant occupational exposures. Apart from vitamin D replacement, she did not take any other medications. She had fully recovered following each episode, but did noticed that she had been becoming more breathless in the preceding 6 months. As part of her workup, blood tests consisting of protein electrophoresis, immunoglobulin profile, aspergillus serology and functional antibodies were sent, all returning reassuringly normal results. Her chest X-ray was unremarkable, as were her lung function tests.

A high-resolution CT of the chest showed equivocal bronchiectatic change at the base of the right lung, and more importantly, a dilated main pulmonary trunk, which was enlarged relative to the ascending aorta (42×32 mm), raising the suspicion of pulmonary hypertension. In light of this, a transthoracic echocardiogram was performed; this showed a dilated right ventricle (RV) measuring 4.4 cm at the base (upper limit of normal 4.2 cm) with preserved function. The left ventricle (LV) was D-shaped throughout the cardiac cycle, suggestive of RV pressure and volume overload. The LV was normal in size with normal systolic and diastolic function. There were no associated significant valvular abnormalities. The patient was then seen in pulmonary hypertension clinic. A nuclear lung ventilation and perfusion scan (VQ) did not show any VQ mismatch, ruling out chronic thromboembolic pulmonary hypertension.

To look for evidence of an intracardiac shunt, a bubble echocardiogram was performed. This revealed a grade 4 (full opacification of the left atrium and LV) shunt during release of Valsalva, which raised the possibility of an interatrial septal defect. The differential diagnosis would have been a patent foramen ovale, though this would not have resulted in significant right heart dilatation For further assessment and in preparation of potential atrial septal defect (ASD) closure, the patient underwent a TOE.

On the morning of the procedure, the patient was oriented to time, place and person. She had a conversation with the operator, and understood all that the procedure entails. She opted to undergo the procedure without sedation, which is offered to all patients in our department. Following the acquisition of consent, her throat was sprayed with topical lidocaine (10%). The patient’s position was then optimised. The procedure was performed with continuous haemodynamic monitoring throughout (in the form of continuous SpO2 monitoring and blood pressure (BP) measurement every 3 min). Intubation of the oesophagus was unremarkable and successful on the first attempt. She remained in sinus rhythm throughout the procedure. Following 20 min of imaging (which confirmed the presence of a haemodynamically significant ASD and, of relevance to our case, a normal left atrial appendage with no evidence of thrombus), the probe was withdrawn clean, with no evidence of blood. Her observations were normal throughout.

Immediately following extubation of the oesophagus, the patient appeared disoriented, and asked ‘where am I?’. The operator allowed the patient to clear her throat and find a comfortable position. The patient remained disoriented, and could not explain how she had travelled to hospital that morning. She was not able to recount the day, month or year. She remained calm throughout.

Investigations

She was immediately transferred to a more appropriate clinical location, where her observations were repeated (BP of 121/95 mm Hg, oxygen saturations of 98% on room air, respiratory rate of 18 breaths/min, temperature of 37.1°C and capillary blood glucose of 5 mmol/L). She exhibited normal tone, power and sensation in her upper and lower limbs. There was no pronator drift. Upper limb (brachial, triceps and brachioradialis) and lower limb (patellar and ankle) reflexes were normal. There was no clonus, dysdiadochokinesia or past-pointing. Cranial nerve examination was unremarkable. The patient was able to read, follow commands and repeat sentences. She had no difficulty in identifying objects. She scored 4/10 in the Abbreviated Mental Test Score (AMTS). The patient appeared to be unable to retain information for longer than 5 min; she asked the same set of questions repeatedly, and did not seem distressed by this.

Differential diagnosis

Given the isolated amnesia in the absence of any neurological physical findings, TGA was top in the list of differential diagnoses. However, in the presence of an ASD and the clear relationship between the TOE and the onset of symptoms, the possibility of a paradoxical embolus causing her symptoms could not be ignored (retching during a TOE is a Valsalva-like manoeuvre, and could theoretically encourage the passage of thrombus from the venous system across the interatrial defect). In view of this, an urgent stroke consultation was sought. Following assessment by the stroke consultant, an MRI of the brain, including diffusion weighted imaging and fluid-attenuated inversion recovery, was performed. This was reviewed by the on-call radiologist and stroke team, both agreeing that there was no evidence of acute infarction. A diagnosis of TGA was, therefore, made.

Treatment

The patient was admitted to the cardiology ward for observation. Her husband was contacted, and on arrival to hospital, he was met by the operator of the procedure. The series of events were recounted, and the diagnosis of TGA was explained. On re-examination the following day, the patient had returned to baseline. She could recall how she arrived to hospital the previous day, and a sensation of gagging. The next thing she could recount is speaking to her daughter in the ward on the evening of the procedure (10 hours after the onset of symptoms).

She scored 10/10 on repeat AMTS. She scored 30/30 on the mini-mental state examination.

Outcome and follow-up

The patient was discharged the following day. Her TOE suggested that her ASD was amenable to percutaneous closure; she is awaiting further assessment by a structural cardiologist.

Discussion

TGA is a clinical syndrome defined by anterograde amnesia in the presence of a normal neurological examination with no localising signs. Often, the clinical feature is the abrupt onset of amnesia; patients suddenly become disoriented in time and ask repetitive questions about the date or their environment (‘broken-record’ phenomenon). Episodes last an average of 6 hours, but may last up to, by definition, 24 hours.¹

The specific symptoms that define TGA suggest that the main site of neurological dysfunction is the mediobasal temporal lobe and hippocampus.² In the acute and postacute stage of TGA, there is a decrease in cerebral blood flow in the temporal lobe and the hippocampal region in most patients, demonstrated in numerous studies where patients underwent single-photon emission computed tomography (SPECT) imaging.³ The aetiological mechanism driving TGA remains unknown; suggested mechanisms include arterial ischaemia (with some studies finding a higher prevalence of vascular risk factors in TGA)⁴ and migraine (with one study demonstrating that patients with TGA were five times more likely to suffer with migraine).⁵ An important recent finding has suggested venous congestion may lead to ischaemia in the previously mentioned structures; numerous studies have shown patients to have a higher incidence of insufficient jugular-vein valves when compared with controls.^6–8 The typical triggers of TGA (physical exercise, sexual intercourse and immersion in cold water) are often associated with Valsalva-like manoeuvres. This would result in increased venous return towards the superior vena cava and, in the presence of jugular venous insufficiency, could lead to ischaemia of the temporal lobe and hippocampus. In our case, intubation of the oesophagus led to a small amount of retching, which may have been enough to, in the presence of jugular venous insufficiency, reduce perfusion to the structures of the brain involved in memory formation. In another case of TGA precipitated by TOE, agitated saline medium was injected at rest and during Valsalva; this was immediately followed by the onset of confusion.⁹ The authors felt that an embolic origin was the likely cause (venous clot crossing the patient’s patent foramen ovale during Valsalva). In our case, we feel this is less likely, as no contrast was injected and no forced Valsalva manoeuvres were performed.

CT and MRI of the brain are often normal or show incidental findings.⁴

Treatment for TGA is not required, with symptoms resolving within 24 hours, as demonstrated in our case.

To the best of our knowledge, only one case of TGA precipitated by TOE (mentioned previously) has been described in the literature, associated with injection of agitated saline (this was not done in our case).⁹

Our case report emphasises the importance of assessment of the patient following any invasive procedure. Prompt clinical evaluation, interspecialty cooperation and an excellent radiology service resulted in prompt assessment and diagnosis in our case.

Patient’s perspective.

I left home on Tuesday, 28 January 2020, at 07:30 to catch the 07:46 train from Theale to Reading for my 08:30 transoesophageal echocardiogram at Royal Berks. I remember buying the ticket, getting on the train. I remember a gagging feeling—it was like a dream. Prior to the appointment, I was concerned about swallowing the probe. The next 9–10 hours are a complete blank. Around late afternoon, I remember talking to my son, then my husband came back after returning home to pick up some essentials for my stay in hospital. I cannot recall much apart from the fact that they were there. My daughter visited about 19:30 (I was not aware of the time). My husband left shortly after my daughter’s arrival. I cannot recall the conversation I had with my son or husband. I can recall some of the conversation I had with my daughter and she left at 21:00. I felt fine the next day. If it was not for the video my husband had recorded of me, I would have found it difficult to comprehend what had happened. It is weird that I cannot recall the day. As far as I know, there are no side effects, I feel great. I am back to normal—playing golf, looking after my granddaughter, working, walking, etc.

Learning points.

• Atypical behaviour following a procedure should elicit prompt examination.

• Good relationship between different specialties is a key in making prompt diagnoses in patients.

• Transient global amnesia can occur following transoesophageal echocardiography, but an acute ischaemic event should be ruled out prior to entertaining this diagnosis.