Convulsive syncope after bidirectional Glenn shunts: physiological implications for a neurological event
http://www.100md.com
《血管的通路杂志》
a Department of Cardiovascular and Thoracic Surgery, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Thiruvananthapuram, Kerala-695011, India
b Department of Cardiology, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Thiruvananthapuram, Kerala-695011, India
Abstract
Objective: Neurological complications after cavopulmonary connections like bidirectional Glenn shunt and Fontan connection are occasionally encountered in the postoperative period. We discuss such a case of bilateral bidirectional Glenn shunt which developed convulsive syncope postoperatively. Case: A 5-year-old cyanotic girl diagnosed as tricuspid atresia with pulmonary stenosis without any spell history underwent bilateral bidirectional Glenn shunt on the way to a subsequent Fontan. After an uneventful surgery she developed convulsive syncope on straining for defecation in the postoperative period. A thorough neurological and arrhythmia study failed to elicit any organic lesions. Discussion: The diagnosis of a neurological event after a single ventricle palliation is paramount to its management. Differentiating syncope from a seizure has its own management implications. The etiologies of neurological complications are varied after cardiac surgery. The physiology and etiology of syncope and seizure after a cavopulmonary connection is discussed. The role of physiological factors in a situation of altered physiodynamics like a bidirectional shunt and Fontan has not been dealt with before in a clinical setting. We have discussed this case to understand the effects of these factors. The effects of strain on the systemic venous pressure, the pulmonary artery pressure and the intrathoracic pressure, can lead to a neurological event if balance is not maintained between the driving pressure of the systemic venous pressure and the pulmonary capacitance. We have devised a simple test to identify these subsets preoperatively by a modification of the Valsalva maneuver. Conclusion: Although neurological complications crop up occasionally after single ventricle palliation, not much in-depth analysis has been done regarding the physiological factors involved after such an altered physiology. The effects of systemic venous pressure, the pulmonary artery pressure and the intrathoracic pressure must be in harmony for proper functioning of the shunt; thus strain can alter physiodynamics to such an extent to manifest clinically as a neurological event. The modified Valsalva maneuver can be applied clinically as a ‘biomarker’ to identify a subset of patients prone for neurological complications.
Key Words: Bi-directional Glenn shunt; Syncope; Seizures; Valsalva maneuver
1. Introduction
Neurological events after palliative procedures like bi-directional Glenn shunt (BDGS, also called superior cavopulmonary anastomosis) and total cavopulmonary connection (TCPC or Fontan procedure) for single ventricle physiology are occasionally encountered in clinical practice [1]. These neurological events may have adverse neurodevelopmental outcomes [2]. The circulatory bypass of the right side of the heart has evolved since Rodbard and Wagner's experimental work on anastomosing the right atrial appendage to the pulmonary artery and ligating the main pulmonary artery in 1949, and Carlon's concept of cavopulmonary shunt in the 1950s, to its first clinical application by Schumacker in 1954 and thereafter its first clinical success by Meshalkin in 1956 [3]. The early successes resulted in aggressive clinical use (or rather abuse), opening the Pandora's Box of unwarranted complications, many of which could have been prevented by strict operative indications [4]. Although the concept of connecting caval blood to the pulmonary circulation caused concern regarding flow across the shunt in moderately elevated pulmonary artery (PA) pressures, the resultant back pressure to the cerebral venous system and the interplay of other physiological factors have not been highlighted enough due to relative lack of data in clinical practice and obviously to the already contraindicated situations due to borderline or high PA pressures. In spite of clinical awareness of neurological problems of a cavopulmonary shunt, a literature review surprisingly revealed few in-depth analyses of the problem [1,5]. Although it can be classified as (a) those predisposed to seizures or stroke attributed to the cyanotic situation per se and (b) those that are related to the operative procedure itself. Still, there are certain occasions which might crop up due to the postoperative status of the patient. Another point which we would like to highlight is the proper diagnosis of a specific neurological event. The diagnosis of syncope or a seizure may be difficult in a child, especially if it is a ‘first event’, and it may be so that one follows the other or vice versa, depending on the pathophysiology of the inciting event [6,7]. The million dollar question remains, which came first, the hen or the egg
We describe a postoperative case of BDGS for a single ventricle physiology – tricuspid atresia with pulmonary stenosis where a neurological event was incited by strain of defecation and which subsequently responded to conservative management. The incidence of ‘strain induced’ syncope and/or seizure or the so-called convulsive syncope had found anecdotal and albeit humouredly passing references in medical journals [8–10]. A study on cough syncope, (one of the situational syncope as just mentioned), demonstrated equalization of arterial and central venous pressures, with concomitant decrease in cerebral blood flow. This led to a rapidly developing cerebral hypoperfusion situation causing syncope [11]. The importance of strain induced syncope is paramount as it is necessary to accurately diagnose a subset of patients which can develop stroke or syncope postoperatively due to strain and to effectively prevent such complications; thus, the way of managing the problem, right from initial diagnosis to the eventual treatment can also be chalked out. We will try to analyze the cause of syncope or seizure and the interplay of physiological mechanisms in the context of a superior cavopulmonary anastomosis. The role of strain induced neurological events after cavopulmonary anastomosis and the way it can affect such patients will also be discussed. Much insight will be put on the clinical diagnosis of the situation and how to deal with it in such a scenario.
2. The present case
We describe a girl with tricuspid atresia and pulmonary stenosis diagnosed at 2 months of age but who was subsequently lost to follow up. She presented to us at 5 years of age with increasing cyanosis. No history of spells or neurological symptoms could be elicited. Examination revealed a cheerful child, who had achieved all milestones but a bit delayed, with deep central cyanosis and clubbing. Her heart rate was regular at 82/min and blood pressure of 106/78 mmHg. Cardiovascular examination revealed a normal S1, single S2 and no murmur. Room air saturations were 68%. Other systemic examinations were within normal limits. Routine investigations showed hemoglobin of 18 gm% with a PCV of 70. Total counts and differential were within normal limits. ESR was 0 at the end of the first hour. Renal and hepatic functions were normal. Chest X-ray showed situs solitus, levocardia, left arch, a CTR of 55% and oligemic lung fields. Transthoracic echocardiography demonstrated situs solitus, levocardia, AV and VA concordance, tricuspid atresia and pulmonary stenosis with peak gradient of 74 mmHg. Bilateral superior vena cavae (SVC) were present and left ventricle was of normal dimensions. An ostium secundum atrial septal defect (ASD) and a 6.5-mm subaortic ventricular septal defect were demonstrated. With these data in hand, the patient was decided for a staged TCPC (bilateral BDGS, followed by TCPC at a later stage as an institutional protocol). Findings at surgery included normally related great vessels, bilateral superior vena cavae, a small patent ductus arteriosus, enlarged and hypertrophied right atrium and a hypoplastic right ventricle. On table PA mean pressure was 16 mmHg. After midline sternotomy, both the SVCs and the branch pulmonary arteries were dissected and looped. Cardiopulmonary bypass was initially instituted with aorto-right atrial cannulation, the PDA ligated and divided. Bilateral BDGS constructed with continuous absorbable suture material. Antegrade flow through the PA was not interrupted as a part of institutional protocol; heart came off bypass at first attempt. Total CPB time was 107 min and core cooled to 32 °C. Surgery was uneventful and the patient was shifted to the ICU for elective ventilation with stable hemodynamics. She was extubated after 8 h of surgery. The post extubation PA pressures stabilized to 14 mmHg and room air saturations were 86%. Forty-eight hours later she was shifted to the wards, subsequently on the 5th postoperative day while straining for defecation, she developed transient loss of consciousness with generalized tonic clonic convulsions which recovered spontaneously within a few seconds. There were no arrhythmias documented during the attack. Neurological evaluation did not reveal any deficits and she went through her recovery well. Next day while passing motions she again had an attack which also had spontaneous recovery. A thorough neurological and cardiological (including Holter monitoring) evaluation failed to reveal any abnormalities. Electroencephalogram showed normal sleep and awake responses. A CT scan done after 72 h of the second attack was normal. The child subsequently was started on stool softeners and the parents were guided for assisting the child with a slow recovery without much strain and physical exercise. The child is on regular follow up with proper guidance regarding her functional status in school and daily activities. She has not reported any similar episodes thereafter.
3. Discussion
3.1. Syncope vs. seizure
The initial problem that any clinician faces with a neurological ‘first event’ in the postoperative cardiac patient is to get to a proper diagnosis. In adult cases a seizure is easily appreciable to syncope than in a pediatric case. More so, altered physiology in palliative shunts like the BDGS in children has a complex pathophysiology along with uncorrected intracardiac shunts. Superimposed to these problems, the child might have congenital cerebral lesions predisposing to seizures. The symptom complex in both syncope and seizure fall in the twilight zone; loss of consciousness, duration, involuntary movements, amnesia and arrhythmias overlap to a variable degree in both [6,7]. The main reason for misdiagnosis between the two is that loss of consciousness is followed sometimes by involuntary movements in syncope, and in epileptic seizures, convulsions almost always precede loss of consciousness; this subtle difference in the presentation is highly observer dependent. The mechanisms of cardiovascular syncope are vasovagal reactions and arrhythmias, whereas in seizure it is due to abnormal electric discharges from neurons. Somewhere interposed between these two symptom complexes is a distinct entity called ‘convulsive syncope’ which is due to neurally mediated vasovagal reactions like carotid sinus hypersensitivity and arrhythmias [7]. We have documented ‘convulsive syncope’ in our patient.
3.2. Etiology of neurological events after cardiac surgery
We propose an etiological classification of neurological complications after cardiac surgery according to the time frame of initiation of events:
Preoperative: (1) Associated neurological problems; (2) Brain abscess due to cyanotic heart disease; (3) Stroke from paradoxical embolism due to intracardiac shunts
Intraoperative: (1) Poor perfusion pressure due to collaterals; (2) Cardiopulmonary bypass complication
Postoperative: (1) Stroke from embolization due to intracardiac shunts; (2) Strain due to altered hemodynamics as in BDGS (our patient); (3) Metabolic
3.3. Pathophysiology of syncope
The basic mechanism of syncope is transient global cerebral hypoperfusion. In a cardiac postoperative case this is often due to a vasovagal reaction which is basically benign or may be a part of response to a hypovolemic state. Otherwise arrhythmias are most often associated with syncope which is the reason why syncope should be evaluated with an ECG rather than a CT scan or EEG. In a patient with BDGS, as in our case, syncope can be either due to strain induced vasovagal reactions, arrhythmias or sudden transient rise in pulmonary artery pressures which cause a functional obstruction to the caval flow to the lungs and subsequently forward flow to the left side. The last point is highly unlikely as compensatory mechanisms will not allow total diminution of forward flow, which, if it occurs, may turn fatal. But it all depends on how much forward flow is maintained. Though ASD may support the systemic circulation, it will not provide adequate pulmonary blood flow for oxygenation leading to transient hypoxia. Straining will also cause a backpressure on the cerebral venous system, but it is transient enough not to cause any prolonged effect. Transient rise can be due to many unknown factors like stress which comes down once the inciting factor is removed. The situation becomes complicated once a seizure activity is noticed. If primary seizure activity is ruled out by the fact that loss of consciousness preceded the seizure, then a diagnosis of ‘convulsive syncope’ can be made clinically. Again this needs to be stressed, that it is basically syncope, and seizure (usually bilateral clonic or myoclonic jerks) is a secondary phenomenon [12]. This needs to be addressed as syncope and not as a seizure. Syncope is usually not associated with a brain lesion. In our patient with a bilateral BDGS, the patient had a convulsive syncope. The absence of any organic lesion, in spite of repeated episodes and the absence of baseline EEG abnormalities, clinched the diagnosis. We will discuss about the factors of strain causing syncope and seizure after a BDGS subsequently.
3.4. Pathophysiology of seizure
Seizure is a manifestation due to paroxysmal discharge of abnormal rhythms in some part of the brain. It is termed epilepsy when these seizures recur, usually spontaneously. Seizure postoperatively is ominous and may in fact be the presenting symptom of severe brain injury, especially if located in a less eloquent area where it is not associated with a deficit. Rarely, arrhythmias can cause a seizure which actually is a convulsive syncope as described earlier. The management of this group of patients is much more complicated and carries a poor prognosis if underlying cerebral insult is a major one, such as in massive infarcts as a result of CPB related complications (hypoperfusion of brain or fat or atheroma shower in adults), or from embolization from an existing intracardiac shunt after a BDGS or TCPC. The avoidance of off-pump BDGS has opened new vistas, where clamping of the venous system may cause transient increase in cerebral venous pressure, but its implications in causing neurological events requires prospective analysis [13].
3.5. Strain induced events in the postoperative period
Broadly, strain usually causes syncope due to a vasovagal response. But there are hardly four reports in medical literature of reflex seizures induced by strains like defecation and even actions like tooth brushing and eating [9,10]. These somatosensory induced seizures are diagnosed with an EEG quite easily. This was ruled out in our patient by the EEG findings. The effects of strain in a patient with a BDGS or TCPC are different and may be augmented relative to a patient with normal hemodynamic connection. The ‘Fontan paradox’ may be a contributory factor [14]. This is one subset of patients which is not much focused on after cardiac surgery especially after BDGS or TCPC. There are primarily three factors in BDGS or TCPC flow dynamics: (1) the systemic venous pressure which is the driving force, (2) the PA pressure which determines the capacitance of the system, i.e. the take up of the blood for the forward flow to the left side (in other words, the primary resistance to the circuit) and (3) the intrathoracic pressure in BDGS or the intraabdominal pressure as in TCPC which causes impediment to the venous flow, the secondary resistance to the circuit. These three factors work in a dynamic way. Each has its role in causing a neurological event. High systemic venous pressures alone, with a normal PA pressure and normal intrathoracic pressure, causes good flow across the shunt and should have an impact on the brain, but this is only a theoretical possibility as clinically, good capacitance and collaterals of the venous system does not allow cerebral venous pressure to rise. Thus, the increase in pressure is well tolerated in superior caval territory compared to inferior caval territory [14]. This may be different in a bilateral shunt as in our case where both of the sides are connected to the pulmonary artery. The PA pressure is the most important singular point for the functioning of the shunt. Any rise in PA pressures, either due to strain or any inciting factors or even an antegrade PA flow, will result in resistance to the passive flow mechanisms thus causing diminution of the forward flow and may be fatal in spite of a shunt being left behind at the atrial level (as in our case). Usually after a BDGS the compensatory right to left shunt across the ASD is not sufficient to maintain left sided flow especially in a slightly older child. The third point is the intrathoracic pressure; this is important in a situation where the child with a shunt may have other factors suitable for BDGS or TCPC. There is a subset of patients who may be prone for strain induced neurological events after a BDGS as in our case. The model of strain which we will follow is the Valsalva maneuver. Whatever be the strain, be it a bout of coughing, or breath holding or defecation (our patient), the VM is the simplest clinical model to comprehend. It has been used as a ‘biomarker’ in many clinical situations, e.g. in evaluating filling pressures in heart failure patients [15]. A combined study on head-up tilt test, transcranial Doppler ultrasonography and VM for diagnosing cerebral syncope demonstrated transient acute complex changes in cardiovascular and cerebrovascular variables. These, along with reduction of cerebrovascular resistance and even paradoxical vasoconstriction, compromise cerebral perfusion pressure and lead to syncope [16]. The response to VM-induced change in ventricular filling in patients with systolic dysfunction as demonstrated by change of a false normal or restrictive left ventricular filling pattern into an abnormal relaxation pattern predicts a lower risk of death or severe heart failure [17]. Variations in technique of the VM have been shown to give variable cardiovascular response in relation to factors like fully expanded lungs and low strain pressure where there is a bradycardic response instead of tachycardia [18]. The unpredictable response may be a deterrent but nevertheless VM remains a useful tool in clinical practice if done classically. By VM we increase the intrathoracic pressure which compromises BDG or TCPC flow dynamics. At the same time, though the systemic venous pressure rises due to the strain, it may not be sufficient to cause forward flow through the pulmonary arteries. The situation is compounded in a bilateral BDGS. In a BDGS the collaterals from the other side, like hemiazygos and other veins act as a vent for decompressing the cerebral venous pressure. Thus, bilateral BDGS may be more prone to neurological complications. The time frame for a guarded recovery is dependent on the increase in the capacitance of the pulmonary vascular bed as the child grows. Incrementing evidence has shown definite growth patterns in central pulmonary arteries following BDGS after 15 months of surgery [14].
3.6. The test
BDGS with antegrade flow may be the final palliation rather than completion TCPC in some of the cases [14,19]. We propose to introduce a modified Valsalva procedure for children who underwent BDGS, to determine these subsets that can develop neurological symptoms in the postoperative period. Any child who can comprehend, is asked to reverse the incentive spirometer (with 3 balls) and blow forcefully with the nose closed and keep it as long as possible (exactly opposite to the respiratory exercises as advised in an incentive spirometer with deep inspiration to lift up the 3 balls and expand the lungs). Simultaneously, the SpO2 probe is connected to note any reduction in room air saturation. Any significant lowering of saturation warns us to guide the parents regarding the level of strain the child may tolerate and to proceed cautiously on their follow up. A quantitative evaluation of the fall in saturation requires further studies in the future. Earlier studies had obtained normal values of pulmonary capillary wedge pressures and blood pressure response to VM in normal elderly subjects which are required to interpret patient data, but its utility in the pediatric single ventricle physiology is doubtful [20].
4. Conclusion
Neurological complications occasionally complicate palliative procedures like BDGS or TCPC for single ventricle physiology. The cause can be related to the preoperative status or intraoperative complications, but also in the postoperative period where residual intracardiac shunts play an important role. But there is a potential situation which had not been described before, that of the altered physiology as a result of the shunt itself. We have described a case of convulsive syncope in a child who underwent a bilateral BDGS as a first stage palliation to a subsequent TCPC. The dynamic interplay of systemic venous pressure (the driving force), the pulmonary artery pressure (the capacitance or the resistance) and the intrathoracic pressure all act in tandem for the optimal functioning of the shunt. As these are all physiological variables, the flow in the shunt may alter at any variations. Contrasting to the normal PA pressures (above 15 mmHg) which are above systemic venous pressures (10 mmHg), in Fontan circulation or for that matter in BDGS, the systemic venous pressures are above or equal to PA pressures [14]. This Fontan paradox may in fact be a contributory factor. In our case, the child had convulsive syncope due to strain during defecation. This gave us an insight to look into the matter from a different angle and we started working on the physiological factors for functioning of the shunt and how this important subset of patients can be detected and prevented from developing any neurological complications in the postoperative period. The introduction of a modified Valsalva test for detecting fall in saturation was a definite step in this direction. But much needs to be done in quantization of the test in terms of percentage fall in saturations and the chance of a neurological event; a prospective study is needed. Further evolutions and newer ideas may be required to detect these subsets at the preoperative stage. A definitive test is required to detect at a much younger child which would be the usual case, as reverse incentive spirometry is not possible in a child <3 years. The Valsalva test should guide us regarding advice for the amount of graded exercises and the guarded step ladder return to normal activities. The time required for attaining near normalcy is dependent on many factors but most important is the growth of the pulmonary arteries after the shunt and the time taken to achieve a growth [21]. Again the long-term implications need to be looked into as aortopulmonary connections, pulmonary arteriovenous fistulae and systemic mediastinal-pulmonary venous collaterals develop late after Fontan surgery due to pressure difference between systemic and pulmonary veins. The bypassing of the hepatic circulation and nonpulsatility of caval blood flow to the pulmonary apices results in failure of recruitment of pulmonary vessels which bring entirely different physiological considerations into perspective [14,22]. However, the role of these shunts in the development of a neurological condition in adult survivors needs further evaluation. In spite of this being a singular case of a neurological event due to the physiological variations, it has given us a model to work on. Similar situations in the future can be confidently avoided if the factors mentioned by us are taken into consideration as we propose to do in these subsets.
References
Kirkham FJ. Recognition and prevention of neurological complications in pediatric cardiac surgery. Pediatr Cardiol 1998; 19:331–345.
Rappaport LA, Wypij D, Bellinger DC, Helmers SL, Holmes GL, Barnes PD, Wernovsky G, Kuban KC, Jonas RA, Newburger JW. Relation of seizures after cardiac surgery in early infancy to neurodevelopmental outcome. Boston Circulatory Arrest Study group. Circulation 1998; 97:773–779.
Konstantinov IE, Alexi-Meskishvili VV. Cavo-pulmonary shunt: from the first experiments to clinical practice. Ann Thorac Surg 1999; 68:1100–1106.
Glenn WWL, Ordway NK, Talner NS, Call EP Jr. Circulatory bypass of the right side of the heart. VI. Shunt between superior vena cava and distal right pulmonary artery; report of clinical application in thirty-eight cases. Circulation 1965; XXXI:172–190.
Reich DL, Griepp RW, Golinko RJ. Severe neurologic injuy following bidirectionl Glenn shunt. J Cardiothorac Vasc Anesth 1993; 7:207–209.
Sheldon R, Rose S, Ritchie D, Connolly D, Koshman ML, Lee MA, Frenneaux M, Fisher M, Murphy W. Historical criteria that distinguish syncope from seizures. J Am Coll Cardiol 2002; 40:142–148.
Begfelt L. Differential diagnosis of cardiogenic syncope and seizure disorders. Heart 2003; 89:353–358.
Braga SS, Manni R, Pedretti RFE. Laughter-induced syncope. Lancet 2005; 366:426.
Harbord MG, Mitchell C. Reflex seizures induced by defecation, with an Ictal EEG focus in the left frontotemporal region. Epilepsia 2002; 43:946–947.
O'Brien TJ, Hogan RE, Sedal L, Murrie V, Cook MJ. Tooth-brushing epilepsy: a report of a case with structural and functional imaging and electrophysiology demonstrating a right frontal focus. Epilepsia 1996; 37:694–697.
Avramov K, Sztriha L, Makai A, Jambrik Z, Rudas L, Vecsei L. Mechanism of cough syncope. Orv Hetil 2004; 145:1625–1627.
Horrocks IA, Nechay A, Stephenson JB, Zuberi SM. Anoxic-epileptic seizures: observational study of epileptic seizures induced by syncopes. Arch Dis Child 2005; 90:183–187.
Luo XJ, Yan J, Wu QY, Yang KM, Xu JP, Liu YL. Clinical application of bidirectional Glenn shunt with off-pump technique. Asian Cardiovasc Thorac Ann 2004; 12:103–106.
Sidi D. Cardiac and pulmonary physiology in the functionally univentricular circulation with reference to total cavopulmonary circulation. Cardiol Young 2005; 15:26–30.
Felker GM, Cuculich PS, Gheorghiade M. The Valsalva maneuver: a bedside ‘biomarker’ for heart failure. Am J Med 2006; 119:117–122.
Chuang YM, Hu HH, Pan PJ. Cerebral syncope: insights from Valsalva maneuver. Eur Neurol 2005; 54:98–102.
Garcia-Lledo A, Moya JL, Balaguer J. Prognostic value of Valsalva maneuver-induced change in Doppler-detected ventricular filling in patients with systolic dysfunction. Rev Esp Cardiol 2005; 58:1029–1036.
Looga R. The Valsalva manoeuvre – cardiovascular effects and performance technique: a critical review. Respir Physiol Neurobiol 2005; 147:39–49.
Lamberti JJ, Spicer RL, Waldman JD, Grehl TM, Thomson D, George L, Kirkpatric SE, Mathewson JW. The bidirectional cavopulmonary shunt. J Thorac Cardiovasc Surg 1990; 100:22–30.
Remmen JJ, Aengevaeren WR, Verheugt FW, Jansen RW. Normal values of pulmonary capillary wedge pressure and the blood pressure response to the Valsalva maneuver in healthy elderly subjects. Clin Physiol Funct Imaging 2005; 25:318–326.
Mendelsohn AM, Bove EL, Lupinetti FM, Crowley DC, Lloyd TR, Beekman RH. Central pulmonary growth patterns after the bidirectional Glenn procedure. J Thorac Cardiovasc Surg 1994; 107:1284–1290.
De Leval MR. The Fontan circulation: a challenge to William Harvey Nat Clin Pract Cardiovasc Med 2005; 2:202–208.(Rajnish Duara, Sumanta Sh)
b Department of Cardiology, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Thiruvananthapuram, Kerala-695011, India
Abstract
Objective: Neurological complications after cavopulmonary connections like bidirectional Glenn shunt and Fontan connection are occasionally encountered in the postoperative period. We discuss such a case of bilateral bidirectional Glenn shunt which developed convulsive syncope postoperatively. Case: A 5-year-old cyanotic girl diagnosed as tricuspid atresia with pulmonary stenosis without any spell history underwent bilateral bidirectional Glenn shunt on the way to a subsequent Fontan. After an uneventful surgery she developed convulsive syncope on straining for defecation in the postoperative period. A thorough neurological and arrhythmia study failed to elicit any organic lesions. Discussion: The diagnosis of a neurological event after a single ventricle palliation is paramount to its management. Differentiating syncope from a seizure has its own management implications. The etiologies of neurological complications are varied after cardiac surgery. The physiology and etiology of syncope and seizure after a cavopulmonary connection is discussed. The role of physiological factors in a situation of altered physiodynamics like a bidirectional shunt and Fontan has not been dealt with before in a clinical setting. We have discussed this case to understand the effects of these factors. The effects of strain on the systemic venous pressure, the pulmonary artery pressure and the intrathoracic pressure, can lead to a neurological event if balance is not maintained between the driving pressure of the systemic venous pressure and the pulmonary capacitance. We have devised a simple test to identify these subsets preoperatively by a modification of the Valsalva maneuver. Conclusion: Although neurological complications crop up occasionally after single ventricle palliation, not much in-depth analysis has been done regarding the physiological factors involved after such an altered physiology. The effects of systemic venous pressure, the pulmonary artery pressure and the intrathoracic pressure must be in harmony for proper functioning of the shunt; thus strain can alter physiodynamics to such an extent to manifest clinically as a neurological event. The modified Valsalva maneuver can be applied clinically as a ‘biomarker’ to identify a subset of patients prone for neurological complications.
Key Words: Bi-directional Glenn shunt; Syncope; Seizures; Valsalva maneuver
1. Introduction
Neurological events after palliative procedures like bi-directional Glenn shunt (BDGS, also called superior cavopulmonary anastomosis) and total cavopulmonary connection (TCPC or Fontan procedure) for single ventricle physiology are occasionally encountered in clinical practice [1]. These neurological events may have adverse neurodevelopmental outcomes [2]. The circulatory bypass of the right side of the heart has evolved since Rodbard and Wagner's experimental work on anastomosing the right atrial appendage to the pulmonary artery and ligating the main pulmonary artery in 1949, and Carlon's concept of cavopulmonary shunt in the 1950s, to its first clinical application by Schumacker in 1954 and thereafter its first clinical success by Meshalkin in 1956 [3]. The early successes resulted in aggressive clinical use (or rather abuse), opening the Pandora's Box of unwarranted complications, many of which could have been prevented by strict operative indications [4]. Although the concept of connecting caval blood to the pulmonary circulation caused concern regarding flow across the shunt in moderately elevated pulmonary artery (PA) pressures, the resultant back pressure to the cerebral venous system and the interplay of other physiological factors have not been highlighted enough due to relative lack of data in clinical practice and obviously to the already contraindicated situations due to borderline or high PA pressures. In spite of clinical awareness of neurological problems of a cavopulmonary shunt, a literature review surprisingly revealed few in-depth analyses of the problem [1,5]. Although it can be classified as (a) those predisposed to seizures or stroke attributed to the cyanotic situation per se and (b) those that are related to the operative procedure itself. Still, there are certain occasions which might crop up due to the postoperative status of the patient. Another point which we would like to highlight is the proper diagnosis of a specific neurological event. The diagnosis of syncope or a seizure may be difficult in a child, especially if it is a ‘first event’, and it may be so that one follows the other or vice versa, depending on the pathophysiology of the inciting event [6,7]. The million dollar question remains, which came first, the hen or the egg
We describe a postoperative case of BDGS for a single ventricle physiology – tricuspid atresia with pulmonary stenosis where a neurological event was incited by strain of defecation and which subsequently responded to conservative management. The incidence of ‘strain induced’ syncope and/or seizure or the so-called convulsive syncope had found anecdotal and albeit humouredly passing references in medical journals [8–10]. A study on cough syncope, (one of the situational syncope as just mentioned), demonstrated equalization of arterial and central venous pressures, with concomitant decrease in cerebral blood flow. This led to a rapidly developing cerebral hypoperfusion situation causing syncope [11]. The importance of strain induced syncope is paramount as it is necessary to accurately diagnose a subset of patients which can develop stroke or syncope postoperatively due to strain and to effectively prevent such complications; thus, the way of managing the problem, right from initial diagnosis to the eventual treatment can also be chalked out. We will try to analyze the cause of syncope or seizure and the interplay of physiological mechanisms in the context of a superior cavopulmonary anastomosis. The role of strain induced neurological events after cavopulmonary anastomosis and the way it can affect such patients will also be discussed. Much insight will be put on the clinical diagnosis of the situation and how to deal with it in such a scenario.
2. The present case
We describe a girl with tricuspid atresia and pulmonary stenosis diagnosed at 2 months of age but who was subsequently lost to follow up. She presented to us at 5 years of age with increasing cyanosis. No history of spells or neurological symptoms could be elicited. Examination revealed a cheerful child, who had achieved all milestones but a bit delayed, with deep central cyanosis and clubbing. Her heart rate was regular at 82/min and blood pressure of 106/78 mmHg. Cardiovascular examination revealed a normal S1, single S2 and no murmur. Room air saturations were 68%. Other systemic examinations were within normal limits. Routine investigations showed hemoglobin of 18 gm% with a PCV of 70. Total counts and differential were within normal limits. ESR was 0 at the end of the first hour. Renal and hepatic functions were normal. Chest X-ray showed situs solitus, levocardia, left arch, a CTR of 55% and oligemic lung fields. Transthoracic echocardiography demonstrated situs solitus, levocardia, AV and VA concordance, tricuspid atresia and pulmonary stenosis with peak gradient of 74 mmHg. Bilateral superior vena cavae (SVC) were present and left ventricle was of normal dimensions. An ostium secundum atrial septal defect (ASD) and a 6.5-mm subaortic ventricular septal defect were demonstrated. With these data in hand, the patient was decided for a staged TCPC (bilateral BDGS, followed by TCPC at a later stage as an institutional protocol). Findings at surgery included normally related great vessels, bilateral superior vena cavae, a small patent ductus arteriosus, enlarged and hypertrophied right atrium and a hypoplastic right ventricle. On table PA mean pressure was 16 mmHg. After midline sternotomy, both the SVCs and the branch pulmonary arteries were dissected and looped. Cardiopulmonary bypass was initially instituted with aorto-right atrial cannulation, the PDA ligated and divided. Bilateral BDGS constructed with continuous absorbable suture material. Antegrade flow through the PA was not interrupted as a part of institutional protocol; heart came off bypass at first attempt. Total CPB time was 107 min and core cooled to 32 °C. Surgery was uneventful and the patient was shifted to the ICU for elective ventilation with stable hemodynamics. She was extubated after 8 h of surgery. The post extubation PA pressures stabilized to 14 mmHg and room air saturations were 86%. Forty-eight hours later she was shifted to the wards, subsequently on the 5th postoperative day while straining for defecation, she developed transient loss of consciousness with generalized tonic clonic convulsions which recovered spontaneously within a few seconds. There were no arrhythmias documented during the attack. Neurological evaluation did not reveal any deficits and she went through her recovery well. Next day while passing motions she again had an attack which also had spontaneous recovery. A thorough neurological and cardiological (including Holter monitoring) evaluation failed to reveal any abnormalities. Electroencephalogram showed normal sleep and awake responses. A CT scan done after 72 h of the second attack was normal. The child subsequently was started on stool softeners and the parents were guided for assisting the child with a slow recovery without much strain and physical exercise. The child is on regular follow up with proper guidance regarding her functional status in school and daily activities. She has not reported any similar episodes thereafter.
3. Discussion
3.1. Syncope vs. seizure
The initial problem that any clinician faces with a neurological ‘first event’ in the postoperative cardiac patient is to get to a proper diagnosis. In adult cases a seizure is easily appreciable to syncope than in a pediatric case. More so, altered physiology in palliative shunts like the BDGS in children has a complex pathophysiology along with uncorrected intracardiac shunts. Superimposed to these problems, the child might have congenital cerebral lesions predisposing to seizures. The symptom complex in both syncope and seizure fall in the twilight zone; loss of consciousness, duration, involuntary movements, amnesia and arrhythmias overlap to a variable degree in both [6,7]. The main reason for misdiagnosis between the two is that loss of consciousness is followed sometimes by involuntary movements in syncope, and in epileptic seizures, convulsions almost always precede loss of consciousness; this subtle difference in the presentation is highly observer dependent. The mechanisms of cardiovascular syncope are vasovagal reactions and arrhythmias, whereas in seizure it is due to abnormal electric discharges from neurons. Somewhere interposed between these two symptom complexes is a distinct entity called ‘convulsive syncope’ which is due to neurally mediated vasovagal reactions like carotid sinus hypersensitivity and arrhythmias [7]. We have documented ‘convulsive syncope’ in our patient.
3.2. Etiology of neurological events after cardiac surgery
We propose an etiological classification of neurological complications after cardiac surgery according to the time frame of initiation of events:
Preoperative: (1) Associated neurological problems; (2) Brain abscess due to cyanotic heart disease; (3) Stroke from paradoxical embolism due to intracardiac shunts
Intraoperative: (1) Poor perfusion pressure due to collaterals; (2) Cardiopulmonary bypass complication
Postoperative: (1) Stroke from embolization due to intracardiac shunts; (2) Strain due to altered hemodynamics as in BDGS (our patient); (3) Metabolic
3.3. Pathophysiology of syncope
The basic mechanism of syncope is transient global cerebral hypoperfusion. In a cardiac postoperative case this is often due to a vasovagal reaction which is basically benign or may be a part of response to a hypovolemic state. Otherwise arrhythmias are most often associated with syncope which is the reason why syncope should be evaluated with an ECG rather than a CT scan or EEG. In a patient with BDGS, as in our case, syncope can be either due to strain induced vasovagal reactions, arrhythmias or sudden transient rise in pulmonary artery pressures which cause a functional obstruction to the caval flow to the lungs and subsequently forward flow to the left side. The last point is highly unlikely as compensatory mechanisms will not allow total diminution of forward flow, which, if it occurs, may turn fatal. But it all depends on how much forward flow is maintained. Though ASD may support the systemic circulation, it will not provide adequate pulmonary blood flow for oxygenation leading to transient hypoxia. Straining will also cause a backpressure on the cerebral venous system, but it is transient enough not to cause any prolonged effect. Transient rise can be due to many unknown factors like stress which comes down once the inciting factor is removed. The situation becomes complicated once a seizure activity is noticed. If primary seizure activity is ruled out by the fact that loss of consciousness preceded the seizure, then a diagnosis of ‘convulsive syncope’ can be made clinically. Again this needs to be stressed, that it is basically syncope, and seizure (usually bilateral clonic or myoclonic jerks) is a secondary phenomenon [12]. This needs to be addressed as syncope and not as a seizure. Syncope is usually not associated with a brain lesion. In our patient with a bilateral BDGS, the patient had a convulsive syncope. The absence of any organic lesion, in spite of repeated episodes and the absence of baseline EEG abnormalities, clinched the diagnosis. We will discuss about the factors of strain causing syncope and seizure after a BDGS subsequently.
3.4. Pathophysiology of seizure
Seizure is a manifestation due to paroxysmal discharge of abnormal rhythms in some part of the brain. It is termed epilepsy when these seizures recur, usually spontaneously. Seizure postoperatively is ominous and may in fact be the presenting symptom of severe brain injury, especially if located in a less eloquent area where it is not associated with a deficit. Rarely, arrhythmias can cause a seizure which actually is a convulsive syncope as described earlier. The management of this group of patients is much more complicated and carries a poor prognosis if underlying cerebral insult is a major one, such as in massive infarcts as a result of CPB related complications (hypoperfusion of brain or fat or atheroma shower in adults), or from embolization from an existing intracardiac shunt after a BDGS or TCPC. The avoidance of off-pump BDGS has opened new vistas, where clamping of the venous system may cause transient increase in cerebral venous pressure, but its implications in causing neurological events requires prospective analysis [13].
3.5. Strain induced events in the postoperative period
Broadly, strain usually causes syncope due to a vasovagal response. But there are hardly four reports in medical literature of reflex seizures induced by strains like defecation and even actions like tooth brushing and eating [9,10]. These somatosensory induced seizures are diagnosed with an EEG quite easily. This was ruled out in our patient by the EEG findings. The effects of strain in a patient with a BDGS or TCPC are different and may be augmented relative to a patient with normal hemodynamic connection. The ‘Fontan paradox’ may be a contributory factor [14]. This is one subset of patients which is not much focused on after cardiac surgery especially after BDGS or TCPC. There are primarily three factors in BDGS or TCPC flow dynamics: (1) the systemic venous pressure which is the driving force, (2) the PA pressure which determines the capacitance of the system, i.e. the take up of the blood for the forward flow to the left side (in other words, the primary resistance to the circuit) and (3) the intrathoracic pressure in BDGS or the intraabdominal pressure as in TCPC which causes impediment to the venous flow, the secondary resistance to the circuit. These three factors work in a dynamic way. Each has its role in causing a neurological event. High systemic venous pressures alone, with a normal PA pressure and normal intrathoracic pressure, causes good flow across the shunt and should have an impact on the brain, but this is only a theoretical possibility as clinically, good capacitance and collaterals of the venous system does not allow cerebral venous pressure to rise. Thus, the increase in pressure is well tolerated in superior caval territory compared to inferior caval territory [14]. This may be different in a bilateral shunt as in our case where both of the sides are connected to the pulmonary artery. The PA pressure is the most important singular point for the functioning of the shunt. Any rise in PA pressures, either due to strain or any inciting factors or even an antegrade PA flow, will result in resistance to the passive flow mechanisms thus causing diminution of the forward flow and may be fatal in spite of a shunt being left behind at the atrial level (as in our case). Usually after a BDGS the compensatory right to left shunt across the ASD is not sufficient to maintain left sided flow especially in a slightly older child. The third point is the intrathoracic pressure; this is important in a situation where the child with a shunt may have other factors suitable for BDGS or TCPC. There is a subset of patients who may be prone for strain induced neurological events after a BDGS as in our case. The model of strain which we will follow is the Valsalva maneuver. Whatever be the strain, be it a bout of coughing, or breath holding or defecation (our patient), the VM is the simplest clinical model to comprehend. It has been used as a ‘biomarker’ in many clinical situations, e.g. in evaluating filling pressures in heart failure patients [15]. A combined study on head-up tilt test, transcranial Doppler ultrasonography and VM for diagnosing cerebral syncope demonstrated transient acute complex changes in cardiovascular and cerebrovascular variables. These, along with reduction of cerebrovascular resistance and even paradoxical vasoconstriction, compromise cerebral perfusion pressure and lead to syncope [16]. The response to VM-induced change in ventricular filling in patients with systolic dysfunction as demonstrated by change of a false normal or restrictive left ventricular filling pattern into an abnormal relaxation pattern predicts a lower risk of death or severe heart failure [17]. Variations in technique of the VM have been shown to give variable cardiovascular response in relation to factors like fully expanded lungs and low strain pressure where there is a bradycardic response instead of tachycardia [18]. The unpredictable response may be a deterrent but nevertheless VM remains a useful tool in clinical practice if done classically. By VM we increase the intrathoracic pressure which compromises BDG or TCPC flow dynamics. At the same time, though the systemic venous pressure rises due to the strain, it may not be sufficient to cause forward flow through the pulmonary arteries. The situation is compounded in a bilateral BDGS. In a BDGS the collaterals from the other side, like hemiazygos and other veins act as a vent for decompressing the cerebral venous pressure. Thus, bilateral BDGS may be more prone to neurological complications. The time frame for a guarded recovery is dependent on the increase in the capacitance of the pulmonary vascular bed as the child grows. Incrementing evidence has shown definite growth patterns in central pulmonary arteries following BDGS after 15 months of surgery [14].
3.6. The test
BDGS with antegrade flow may be the final palliation rather than completion TCPC in some of the cases [14,19]. We propose to introduce a modified Valsalva procedure for children who underwent BDGS, to determine these subsets that can develop neurological symptoms in the postoperative period. Any child who can comprehend, is asked to reverse the incentive spirometer (with 3 balls) and blow forcefully with the nose closed and keep it as long as possible (exactly opposite to the respiratory exercises as advised in an incentive spirometer with deep inspiration to lift up the 3 balls and expand the lungs). Simultaneously, the SpO2 probe is connected to note any reduction in room air saturation. Any significant lowering of saturation warns us to guide the parents regarding the level of strain the child may tolerate and to proceed cautiously on their follow up. A quantitative evaluation of the fall in saturation requires further studies in the future. Earlier studies had obtained normal values of pulmonary capillary wedge pressures and blood pressure response to VM in normal elderly subjects which are required to interpret patient data, but its utility in the pediatric single ventricle physiology is doubtful [20].
4. Conclusion
Neurological complications occasionally complicate palliative procedures like BDGS or TCPC for single ventricle physiology. The cause can be related to the preoperative status or intraoperative complications, but also in the postoperative period where residual intracardiac shunts play an important role. But there is a potential situation which had not been described before, that of the altered physiology as a result of the shunt itself. We have described a case of convulsive syncope in a child who underwent a bilateral BDGS as a first stage palliation to a subsequent TCPC. The dynamic interplay of systemic venous pressure (the driving force), the pulmonary artery pressure (the capacitance or the resistance) and the intrathoracic pressure all act in tandem for the optimal functioning of the shunt. As these are all physiological variables, the flow in the shunt may alter at any variations. Contrasting to the normal PA pressures (above 15 mmHg) which are above systemic venous pressures (10 mmHg), in Fontan circulation or for that matter in BDGS, the systemic venous pressures are above or equal to PA pressures [14]. This Fontan paradox may in fact be a contributory factor. In our case, the child had convulsive syncope due to strain during defecation. This gave us an insight to look into the matter from a different angle and we started working on the physiological factors for functioning of the shunt and how this important subset of patients can be detected and prevented from developing any neurological complications in the postoperative period. The introduction of a modified Valsalva test for detecting fall in saturation was a definite step in this direction. But much needs to be done in quantization of the test in terms of percentage fall in saturations and the chance of a neurological event; a prospective study is needed. Further evolutions and newer ideas may be required to detect these subsets at the preoperative stage. A definitive test is required to detect at a much younger child which would be the usual case, as reverse incentive spirometry is not possible in a child <3 years. The Valsalva test should guide us regarding advice for the amount of graded exercises and the guarded step ladder return to normal activities. The time required for attaining near normalcy is dependent on many factors but most important is the growth of the pulmonary arteries after the shunt and the time taken to achieve a growth [21]. Again the long-term implications need to be looked into as aortopulmonary connections, pulmonary arteriovenous fistulae and systemic mediastinal-pulmonary venous collaterals develop late after Fontan surgery due to pressure difference between systemic and pulmonary veins. The bypassing of the hepatic circulation and nonpulsatility of caval blood flow to the pulmonary apices results in failure of recruitment of pulmonary vessels which bring entirely different physiological considerations into perspective [14,22]. However, the role of these shunts in the development of a neurological condition in adult survivors needs further evaluation. In spite of this being a singular case of a neurological event due to the physiological variations, it has given us a model to work on. Similar situations in the future can be confidently avoided if the factors mentioned by us are taken into consideration as we propose to do in these subsets.
References
Kirkham FJ. Recognition and prevention of neurological complications in pediatric cardiac surgery. Pediatr Cardiol 1998; 19:331–345.
Rappaport LA, Wypij D, Bellinger DC, Helmers SL, Holmes GL, Barnes PD, Wernovsky G, Kuban KC, Jonas RA, Newburger JW. Relation of seizures after cardiac surgery in early infancy to neurodevelopmental outcome. Boston Circulatory Arrest Study group. Circulation 1998; 97:773–779.
Konstantinov IE, Alexi-Meskishvili VV. Cavo-pulmonary shunt: from the first experiments to clinical practice. Ann Thorac Surg 1999; 68:1100–1106.
Glenn WWL, Ordway NK, Talner NS, Call EP Jr. Circulatory bypass of the right side of the heart. VI. Shunt between superior vena cava and distal right pulmonary artery; report of clinical application in thirty-eight cases. Circulation 1965; XXXI:172–190.
Reich DL, Griepp RW, Golinko RJ. Severe neurologic injuy following bidirectionl Glenn shunt. J Cardiothorac Vasc Anesth 1993; 7:207–209.
Sheldon R, Rose S, Ritchie D, Connolly D, Koshman ML, Lee MA, Frenneaux M, Fisher M, Murphy W. Historical criteria that distinguish syncope from seizures. J Am Coll Cardiol 2002; 40:142–148.
Begfelt L. Differential diagnosis of cardiogenic syncope and seizure disorders. Heart 2003; 89:353–358.
Braga SS, Manni R, Pedretti RFE. Laughter-induced syncope. Lancet 2005; 366:426.
Harbord MG, Mitchell C. Reflex seizures induced by defecation, with an Ictal EEG focus in the left frontotemporal region. Epilepsia 2002; 43:946–947.
O'Brien TJ, Hogan RE, Sedal L, Murrie V, Cook MJ. Tooth-brushing epilepsy: a report of a case with structural and functional imaging and electrophysiology demonstrating a right frontal focus. Epilepsia 1996; 37:694–697.
Avramov K, Sztriha L, Makai A, Jambrik Z, Rudas L, Vecsei L. Mechanism of cough syncope. Orv Hetil 2004; 145:1625–1627.
Horrocks IA, Nechay A, Stephenson JB, Zuberi SM. Anoxic-epileptic seizures: observational study of epileptic seizures induced by syncopes. Arch Dis Child 2005; 90:183–187.
Luo XJ, Yan J, Wu QY, Yang KM, Xu JP, Liu YL. Clinical application of bidirectional Glenn shunt with off-pump technique. Asian Cardiovasc Thorac Ann 2004; 12:103–106.
Sidi D. Cardiac and pulmonary physiology in the functionally univentricular circulation with reference to total cavopulmonary circulation. Cardiol Young 2005; 15:26–30.
Felker GM, Cuculich PS, Gheorghiade M. The Valsalva maneuver: a bedside ‘biomarker’ for heart failure. Am J Med 2006; 119:117–122.
Chuang YM, Hu HH, Pan PJ. Cerebral syncope: insights from Valsalva maneuver. Eur Neurol 2005; 54:98–102.
Garcia-Lledo A, Moya JL, Balaguer J. Prognostic value of Valsalva maneuver-induced change in Doppler-detected ventricular filling in patients with systolic dysfunction. Rev Esp Cardiol 2005; 58:1029–1036.
Looga R. The Valsalva manoeuvre – cardiovascular effects and performance technique: a critical review. Respir Physiol Neurobiol 2005; 147:39–49.
Lamberti JJ, Spicer RL, Waldman JD, Grehl TM, Thomson D, George L, Kirkpatric SE, Mathewson JW. The bidirectional cavopulmonary shunt. J Thorac Cardiovasc Surg 1990; 100:22–30.
Remmen JJ, Aengevaeren WR, Verheugt FW, Jansen RW. Normal values of pulmonary capillary wedge pressure and the blood pressure response to the Valsalva maneuver in healthy elderly subjects. Clin Physiol Funct Imaging 2005; 25:318–326.
Mendelsohn AM, Bove EL, Lupinetti FM, Crowley DC, Lloyd TR, Beekman RH. Central pulmonary growth patterns after the bidirectional Glenn procedure. J Thorac Cardiovasc Surg 1994; 107:1284–1290.
De Leval MR. The Fontan circulation: a challenge to William Harvey Nat Clin Pract Cardiovasc Med 2005; 2:202–208.(Rajnish Duara, Sumanta Sh)