Pathophysiology of Neurally Mediated Syncope: Role of Cardiac Output and Total Peripheral Resistance

Abstract

Syncope is a common clinical condition occurring even in otherwise healthy people without underlying cardiovascular disease. Neurally mediated syncope is by far the the most common cause of syncope in individuals without any structural heart disease. Based on traditional wisdom, loss of sympathetic tone with relaxation of vascular smooth muscle is the key mechanism underlying the pathophysiology of syncope, especially in patients without an acute decrease in heart rate. However, this concept has recently been challenged. Some microneurographic studies indicate that sympathetic withdrawal may not always be a prerequisite even for the development of classic “vasodepressor” forms of syncope. Conversely, a decrease in cardiac output appears to be a determinant factor for syncope in most circumstances. This article reviews the relative contribution of cardiac output versus sympathetic vasoconstriction in neurally mediated syncope in otherwise healthy individuals. It is suggested that a moderate to severe fall in cardiac output with or without vasodilatation may contribute to syncope.

Introduction

Over 1 million people in the United States (Soteriades et al., 2002; Malasana et al., 2011), many of whom are young women (Robertson, 1999), suffer from syncope each year. Neurally mediated (reflex) syncope is by far the most common cause of syncope in people without structural heart disease, but the exact mechanisms for these neural pathways and their variations among individuals remain unclear. Many drugs have been tested in the treatment of recurrent syncope, for the most part with disappointing outcomes (Moya et al., 2009). One major reason for the management challenge may be the complexity of the hemodynamic characteristics of syncope.

Based on the Sharpey-Schafer model (Sharpey-Schafer, 1956), syncope is produced by acute systemic vasodilatation, which may be elicited by forceful contractions of an empty left ventricle, the so called “empty heart syndrome”. Profound vasodilatation in the forearm has indeed been observed during syncope (Barcroft et al., 1944; Barcroft & Edholm, 1945; Epstein et al., 1968; Halliwill et al., 1996; Dietz et al., 1997a; Dietz et al., 1997b). It has been thought that loss of sympathetic tone (Wallin & Sundlof, 1982; Morillo et al., 1997; Mosqueda-Garcia et al., 1997) leading to relaxation of vascular smooth muscle (Barcroft et al., 1944; Barcroft & Edholm, 1945; Sharpey-Schafer, 1956; Epstein et al., 1968) contributes importantly to neurally mediated syncope. However, this traditional concept has recently been challenged.

The new concept is that mechanisms other than sympathetic neural control may be responsible for the initiation and occurrence of syncope. This notion seems to be supported by several microneurographic studies showing that withdrawal of sympathetic activity is not always a prerequisite for (pre)syncope despite significant decreases of arterial pressure (Cooke et al., 2009; Vaddadi et al., 2010; Fu et al., 2012).

Neurally mediated syncope is traditionally defined by three subtypes; vasodepressor, cardioinhibitory, and mixed type (Moya et al., 2009). Recently, we have identified more precise characteristics of syncope based on the contribution of cardiac output and sympathetic vasoconstriction to hypotension (Fu et al., 2012). Our data show that a moderate fall in cardiac output with coincident vasodilatation occurs in the majority (64%) of presyncopal individuals, while a severe fall in cardiac output, driven predominantly by a decrease in heart rate with no changes in total peripheral resistance at presyncope occurs in a smaller (36%) subset.

Contribution of cardiac output to syncope

Results from several laboratories suggest that cardiac output may be a key factor for syncope in humans. For example, Jardine et al reported, for the first time, that a progressive decrease in cardiac output may contribute to hypotension some minutes before syncope occurs (Jardine et al., 2002). This work was furthered by studies showing that sublingual nitroglycerin may facilitate syncope by reducing cardiac output without evidence of sympathetic inhibition (Gisolf et al., 2004). Subsequent research showed that circulatory patterns precipitating presyncope were not different between patients who used versus who did not use sublingual nitroglycerin (Verheyden et al., 2008). In both groups marked hypotension was cardiac output-mediated, suggesting the underlying mechanism may be independent of the use of nitroglycerin (Verheyden et al., 2008). Recently, we observed that all presyncopal individuals had a moderate to severe fall in cardiac output 1-2 minutes prior to syncope (Fu et al., 2012). The fall in cardiac output can be driven by a decrease in heart rate (bradycardia) and/or a decrease in stroke volume.

Profound bradycardia or asystole of course will lead to a marked reduction in cardiac output. This notion was supported by our findings showing that the severe fall in cardiac output prior to syncope was driven predominantly by the decrease of heart rate (Fu et al., 2012). However, previous atropine and cardiac pacing studies have yielded conflicting results, indicating that bradycardia may not always be the only factor in restricting cardiac output (Weissler et al., 1957; Fujimura et al., 1989; Sra et al., 1993; Connolly et al., 1999; Sutton et al., 2000; Ammirati et al., 2001; Connolly et al., 2003; Raviele et al., 2004). The timing of atropine injection may be critical; once presyncope begins, it is probably too late to stop it. Conversely, it is important to select patients who are good candidates for cardiac pacing. The International Study on Syncope of Unknown Etiology (ISSUE) 2 study hypothesized that spontaneous asystole should form the basis for patient selection for pacemaker therapy (Brignole et al., 2006). Indeed, the recently published double-blind, randomized placebo-controlled ISSUE 3 trial has shown that permanent pacing is effective in reducing recurrence of syncope in patients ≥40 years with severe asystolic neurally mediated syncope (Sutton et al., 2014). These results further support the notion that profound bradycardia or asystole plays a crucial role in the fall of cardiac output and the occurrence of syncope.

Many investigations have found that stroke volume decreases progressively and then rapidly at syncope (Jardine et al., 2002; Kozlowski et al., 2003; Verheyden et al., 2007; Verheyden et al., 2008; Folino et al., 2010a; Fu et al., 2012). Excessive blood pooling in the lower body and reduced venous return are responsible for the decrease in stroke volume during orthostasis. It was found that in patients with unexplained recurrent syncope, calf volume changes did not correlate to a symptomatic outcome during upright tilt (Bellard et al., 2003). In addition, graded calf compression stockings failed to improve orthostatic tolerance in healthy individuals (Protheroe et al., 2011). These findings suggest that blood pooling in the splanchnic and/or renal vascular bed(s) rather than the calf may contribute more importantly to syncope (Stewart et al., 2004; Jardine, 2013). However, we cannot exclude the possibility that more blood may be pooled above the calf during orthostasis. The rapid decrease in stroke volume at syncope may be attributed to a reduction of cardiac contractility (Folino et al., 2010b) or acute (rapid) venodilatation (Manyari et al., 1996; Thomson et al., 1996). An earlier study showed that venoconstriction rather than venodilatation occurred in the forearm or hand veins during syncope (Epstein et al., 1968). It has been suggested that active venoconstriction in the limbs is not important to mild orthostatic response (Stewart et al., 2001). Perhaps the problem is still rapid arteriolar vasodilatation, leading to redistribution of cardiac output even without venodilation at the onset of syncope.

Role of sympathetic activity and total peripheral resistance in syncope

Recent studies indicate that withdrawal of sympathetic activity may not directly contribute to syncope (Cooke et al., 2009; Vaddadi et al., 2010; Fu et al., 2012). For example, it was found that sympathetic withdrawal occurred rapidly at presyncope, much later than the initial decrease in blood pressure (Cooke et al., 2009; Fu et al., 2012; Schwartz et al., 2013). Persistence of muscle sympathetic nerve activity (MSNA) was reported during syncope in some people (Vaddadi et al., 2010). We have also observed that in some individuals (all young females) sympathetic withdrawal does not occur during presyncope, and these individuals have a profound increase in plasma epinephrine concentration at presyncope compared with those who show sympathetic withdrawal. These results raise the possibility that activation of the β₂-adrenergic receptor without sympathetic inhibition may also contribute to syncope. Indeed, Hart et al have demonstrated that the β-adrenergic receptors offset α-adrenergic vasoconstriction in young women but not young men or postmenopausal women (Hart et al., 2011). Whether β₂-adrenergically mediated syncope occurs exclusively in young women needs to be investigated.

It is important to note that MSNA recordings are inherently limited to the skeletal muscle vasculature. However, animal studies showed that under most physiological conditions MSNA correlates and coheres approximately with renal and cardiac sympathetic nerve activities (Kamiya et al., 2006). Whether sympathetic withdrawal in other vascular beds follows the same pattern as that in the skeletal muscle during syncope in humans remains unknown. It is also important to note that MSNA analyzed using the conventional methods may not reflect actual changes in sympathetic activity since it is an integrated signal. Analyzing nerve signal before integration, such as raw action potential, firing properties, and sympathetic neural recruitment patterns (Salmanpour et al., 2008; Salmanpour et al., 2011) may provide more insight into the role of sympathetic neural control in syncope.

We found that in the majority of presyncopal individuals, total peripheral resistance decreases earlier than the obvious reduction of MSNA (Fu et al., 2012). However, we cannot exclude the possibility that vasodilatation in other vascular beds, such as splanchnic and renal circulations, could occur early in the development of syncope. It also is possible that the transduction of sympathetic outflow into peripheral vascular resistance may be altered before sympathetic withdrawal actually occurs (Iwase et al., 2002; Kamiya et al., 2005; Cooke et al., 2009). Additionally, there may be a mismatch between nerve firing and norepinephrine release (Vaddadi et al., 2011). Finally, β₂-adrenergic receptor mediated and/or humorally mediated active vasodilatation may contribute to the early decrease in total peripheral resistance (Dietz et al., 1997b; Goldstein et al., 2003). In our study, those with a prominent fall in cardiac output due to a fall in heart rate (all young females), total peripheral resistance remained stable prior to and during presyncope despite the rapid decrease in MSNA (Fu et al., 2012). These individuals may have excessive vasoconstriction in other vascular beds rather than the skeletal muscle. Alternatively, the decrease in MSNA does not lead to a corresponding decrease in total peripheral resistance in them.

In summary, a moderate fall in cardiac output with coincident vasodilatation, or a severe fall in cardiac output without changes in total peripheral resistance contributes importantly to syncope. Sympathetic withdrawal may not always be a prerequisite for neurally mediated syncope. Of note, MSNA analyzed using the conventional methods may not reflect actual changes in sympathetic activity since it is an integrated signal. Analyzing raw nerve signal before integration is needed to confirm these findings.