Although it is now widely accepted that the key to accurate diagnosis and risk stratification of syncope is a thoughtful and scrupulous history, exactly what is meant by the history remains unclear, and moving it from experts to front-line workers has proven difficult. Partly this is because syncope is simply a symptom, like fever, with a plethora of potential causes. Partly as well this reflects the multitude of somewhat overlapping symptoms and signs for the most common form, the ‘faint’ or vasovagal syncope. There are several clinical features that are known to be helpful in the differential diagnosis of loss of consciousnesses, based on quantitative symptom studies. These for example help distinguish epileptic convulsions and pseudosyncope from syncope, but such studies were aimed at clinical decision-making. They reported just the most highly significant clinical points and do not help clinicians make sense of the welter of symptoms that clinical experience suggests (Sheldon et al., 2002, 2006; Wieling et al., 2009; Tannemaat et al., 2013).
In this issue of Brain, van Dijk et al. (2014) provide fascinating and informative insights into why some symptoms cluster with each other in patients with vasovagal syncope. The authors performed detailed EEG and videometric analyses of 69 patients with positive responses to head-up tilt table testing. The EEG provides an objective marker of brain dysfunction during the cerebral hypoperfusion that accompanies syncope. For nearly 60 years investigators have described EEG patterns during provoked reflex syncope (Gastaut and Fischer-Williams, 1957; Ammirati et al., 1998; Sheldon et al., 1998; Martinez-Fernandez et al., 2008). Two patterns have been described. A ‘slow-flat-slow’ pattern is characterized by an initial slow phase in which delta waves appear and wave amplitude increases, a sudden flattening of the EEG, and a return to normal brain activity through a slow phase. A ‘slow’ pattern consists only of an increasing and decreasing slowing. The slow-flat-slow pattern was considered to be a sign of more severe cerebral hypoperfusion and has been observed more commonly in cardio-inhibitory syncope compared with the other subtypes, and associated with the presence of convulsive movements during syncope. However, with the exception of small studies and the specific evaluation of convulsive-like movements during syncope, few studies have investigated both EEG patterns and signs and symptoms during syncope.
Van Dijk et al. (2014) describe signs and haemodynamic and EEG changes using videometric data in 69 cases of tilt-induced vasovagal syncope. Their results confirm that EEG flattening is a sign of more severe cerebral hypoperfusion than EEG slowing alone. Indeed, patients with the slow-flat-slow EEG pattern had a lower minimum blood pressure, longer ECG pauses and duration of loss of consciousness and more often had asystole than those patients with the slow EEG pattern. In addition, the authors analysed several clinical signs occurring during syncope, according to the EEG pattern of the patient and the phase in which they occurred. As already underlined in previous studies, pallor, sweating and having the eyes open during syncope occurred in almost every patient: the five patients with eyes closed during syncope were all in the slow pattern group, suggesting that eye opening might not occur if the brain hypoperfusion is mild. Dilated pupils, making sounds, jerks, eye movements and oral automatisms were observed in >50% of the population. The authors propose a classification of clinical signs based on their relation to the EEG phase. Type A signs occur during the first slow phase, are present during the flat phase and stop in the second slow phase. They include loss of consciousness, eye opening and general stiffening with possible arm raising. Type B signs occur only during EEG slowing and they likely require some cortical involvement. They include myoclonic jerks, nonsensical talking, non-verbal sounds, and automatisms. Patients also reported auras and near-death experiences. Type C signs occur only during EEG flattening and are mainly roving eye movements and stertorous breathing, suggestive of subcortical generation. Finally, type D signs, including jaw dropping and snoring, are less specific and may occur in either phase.
The van Dijk et al. (2014) paper provides an important contribution to our understanding of syncope symptoms. It offers a unique opportunity to analyse reflex syncope symptoms and signs related to EEG changes, and therefore to clarify the pathophysiology of some of the clinical events related to syncope. It provides insight into why some symptoms are clustered with each other. Moreover, signs that have never been described as syncope-related could now help clinicians to distinguish syncope from other causes of loss of consciousness.
Nevertheless, before being able to extend the study’s findings to everyday clinical practice, we must consider that the study has a few weaknesses. First, a very selected population in a highly monitored setting was analysed. The same findings may not be true in either spontaneous vasovagal syncope or syncope from other causes. For example, the onset of partial seizures has been observed in syncope associated with profound bradyarrhythmia or asystole during implantable loop recordings (Petkar et al., 2012). Therefore, in the absence of control groups, such as patients with non-syncopal loss of consciousness and with syncope from other causes, it may be a while before we are confident that these clinical sign patterns are useful in the differential diagnosis of transient loss of consciousness. Moreover, signs that can easily be observed after intensive review of a video might not be easily recalled and described by a bystander after a ‘real life’ syncope episode. Second, the video analysis might be biased by both the absence of blinding to EEG results and the camera position, which did not allow the analysis of peripheral movements.
In conclusion, from a clinical standpoint, this study can help us in recognizing syncope associated with a deeper hypoperfusion, which might be more symptomatic and require more diagnostic and therapeutic effort. Moreover, the description of symptoms that have rarely been described as associated with syncope, such as oral automatisms, might eventually help with the differential diagnosis of transient loss of consciousnesses with atypical clinical patterns.