Anaesthesia for Caesarean section in patients with aortic stenosis: the case for general anaesthesia

A decreasing incidence of rheumatic heart disease and the advances in medical and surgical treatment have resulted in an increase in women with congenital heart disease presenting in pregnancy. Since 1961 only three out of the 89 deaths from congenital heart disease as reported in the Confidential Enquiries into Maternal Deaths [1, 2] have occurred in women with congenital aortic stenosis and none of these was related to anaesthesia. Evidence for the choice of anaesthetic technique is therefore based on clinical experience and on the underlying functional disorder. Cardiovascular pressures, flows and electrical activity can be measured and there is a strong case for establishing guidelines for basic cardiovascular monitoring so that techniques for rare clinical cases can be compared more easily. Overall cardiovascular function is a product of structural abnormalities, pathophysiology and therapeutic interventions. What then are the functional limitations of regional techniques? It is assumed that the use of regional anaesthesia in a parturient with aortic stenosis reported later in this issue [3] is in a woman with the congenital rather than rheumatic variety of aortic stenosis. This is usually the result of a bicuspid aortic valve with the coronary ostia normally placed and various degrees of thickening and fusion of the commissures leading to valvular, supravalvular, membranous or tunnel type of subvalvular stenosis. These abnormalities cause left ventricular outflow obstruction and the degree of risk depends mainly on the severity of the obstruction. The size of the valve opening will control flow and pressure changes across the valve. The critical valve area is <0.6 cm2 for severe stenosis [4]. If the pressure gradient across the valve is estimated by echocardiography in pregnancy, derived values may lead to errors because stroke volume is increased and higher pressure gradients occur; conversely when the heart starts to fail a modest pressure gradient can be recorded leading to a false sense of security. Firstly, maintenance of coronary blood flow must be considered. During pregnancy cardiac output normally increases and this increase in left ventricular work demands augmentation of coronary blood flow. In aortic stenosis the coronary blood vessels are distal to the obstruction and are supplied with blood mainly during diastole; in the pregnant state it becomes more difficult to maintain adequate blood flow to the left ventricle because of increasing systolic and end diastolic ventricular pressures. The work of the ventricle is increased and thus it requires a greater coronary artery blood flow; as this cannot be achieved the patient may experience angina and suffer subendocardial ischaemia. Regional anaesthesia has been associated with ECG changes of ischaemia of multifactorial origin in healthy parturients [5]. The detection of transient ischaemias requires capture of ECG data from leads II and V5 and an analysis of changes from control. During regional anaesthesia in pregnant women, when decreases in systemic vascular resistance can compromise myocardial blood flow detection of ischaemia by this type of ECG, monitoring may be difficult, because of patient movement. In spite of the spinal microcatheter technique used in the case report in this journal, colloid infusion and ephedrine 15 mg intravenously, the patient's systolic blood pressure decreased to 70 mmHg just before delivery. The authors claim that there were no significant changes in haemodynamic parameters; however, it is the diastolic blood pressure which determines myocardial blood flow and this is not stated. Secondly, the reduction in preload induced by regional block is a limitation of the technique. The normal pregnant woman compensates for a decrease in systemic vascular resistance by increasing stroke volume and heart rate to maintain cardiac output. The patient with aortic stenosis has a fixed stroke volume and to maintain cardiac output must elevate her heart rate, but this compromises left ventricular filling. The pregnant woman with aortic stenosis is extremely intolerant of change in left ventricular preload. A decrease in preload caused by haemorrhage or associated with regional anaesthesia can produce cardiogenic shock. An increase in preload can precipitate pulmonary oedema. These are acute changes, but they may complicate a more chronic left ventricular hypertrophy which ultimately progresses to congestive cardiac failure. Control over left ventricular preload is less precise with regional techniques because ventilation cannot be manipulated as would be possible with IPPV and filling pressures are less predictable because they depend on fluid load and altered sympathetic nervous system responses. A segmental nerve block from the lowest sacral segment to T4 is necessary by any route to ensure adequate pain relief during Caesarean section and this invariably produces extensive sympathetic block [6]. Moreover, the risk of hypotension cannot be eliminated. Thirdly, obstetric anaesthetists are agreed that in severe aortic stenosis tachycardia must be prevented. This is confirmed by a case report from the USA which describes the administration of epidural analgesia with full cardiovascular monitoring to a multiparous woman in labour with a rheumatic severe regurgitant and moderately stenotic mitral valve lesion complicated by atrial fibrillation [7]. The authors correctly identify that had this been a severe stenotic lesion a tachycardia could have precipitated pulmonary oedema. Measurement of central venous pressure alone would not detect this event until after it occurred and the authors recommend pulmonary artery wedge pressure measurements in such a severe lesion so that left atrial preload can be assessed. One of the complications of pulmonary artery catheterisation occurred in this case, that of right bundle branch block. Tachycardia is well tolerated in regurgitant valve disease but can lead to a marked increase in left atrial pressure in mitral or aortic stenosis, or in severe cases of hypertrophic cardiomyopathy. Apart from physiological changes in cardiac function which have to be considered in any technique, adverse therapeutic events must also be anticipated. If ephedrine is chosen as the vasoconstrictor, it has α and β effects with a resultant tachycardia which is undesirable in a patient with aortic stenosis. Phenylephrine might be a better choice although it is a pure α agonist [8]. After considering the pathophysiological evidence which would support the use of general anaesthesia in preference to regional nerve block with sympathetic blockade because of the ability to maintain coronary blood flow with adequate filling volumes and diastolic pressures, the next question is, what is the clinical evidence to support one's choice of technique? Recent reviews [9–11] of anaesthesia for the delivery of parturients with aortic stenosis focus on the method of delivery, because a Caesarean section is only indicated for obstetric reasons or because of a sudden deterioration in the patient's condition. During any delivery, changes in the circulation occur, regardless of the anaesthetic used. Blood loss reduces blood volume and to counteract this the uterus contracts and releases additional blood into the circulation. This process is augmented by the use of oxytocin after delivery to ensure uterine contraction and prevent postpartum haemorrhage. One of the deaths reported in the maternal mortality reports describes postpartum cardiac failure exacerbated by oxytocin in a woman with aortic stenosis who required manual removal of the placenta [12]. Unfortunately, details of the anaesthetic used are not given, but there is a time when conversion to general anaesthesia has to be made because of clinical deterioration. When and how this choice is made should be an elective decision rather than in extremis. Generally the choice lies between a regional technique and general anaesthesia, although Arias & Pineda [13] in their case report chose a field block. Regional and general anaesthesia both have their advocates and in a clinical disorder where few patients present for anaesthesia, it is not possible to assess risks based on randomised controlled trials. Each patient must be serially assessed during pregnancy by cardiological investigations, including noninvasive Doppler echocardiography and in some cases cardiac catheterisation. Then, using case reports and professional recommendations, the advantages versus the disadvantages of the particular techniques have to be weighed. The UK 1996 recommendations by the Royal College of Physicians [14] advise standardisation of investigations and the process of letting symptoms guide management has become no longer acceptable. The report has implications for anaesthetic obstetric management because it specifically states that regional anaesthesia is contraindicated for severe aortic stenosis. The case for general anaesthesia is made on the basis that the avoidance of sympathetic blockade which occurs with regional anaesthesia decreases the risk of significant hypotension following a reduction in systemic vascular resistance. In pregnancy hypotension will compromise not only the maternal myocardium but also the placental blood flow to the fetus. One disadvantage of general anaesthesia is the sympathetic nervous system response to intubation, which can generate tachycardia and hypertension, leading to sudden fluctuations in cardiac output. This can be controlled by induction with a cardiostable drug followed by a short acting opioid (e.g. alfentanil), as described in the case report of a woman with severe aortic stenosis presenting for Caesarean section [15]. Volatile anaesthetic agents also have a direct myocardial depressant effect but this is dose related. In obstetrics, their concentration is limited because of their relaxant effect on the myometrium. Rarely, the sensory nerve blockade is not fully effective and resort to general anaesthesia must be made. During any regional technique in a parturient with cardiac disease a decision must be taken as to when to convert to a general anaesthetic, for example if desaturation occurs or if the mother becomes breathless. Adequate monitoring is vital to enable a decision to be made without delay. Essential monitoring techniques include continuous ECG signals from leads II and V5 to detect endocardial ischaemia. Central venous pressure recordings will allow right atrial pressure monitoring. In addition, an indwelling intra-arterial catheter will allow direct arterial monitoring and sampling of the blood gas and acid–base status of the mother. Cardiac output can also be monitored by a number of direct or indirect methods. A pulmonary artery catheter allows measurement of pulmonary artery and pulmonary wedge pressures, reflecting left atrial pressure and left ventricular filling, the maintenance of which is critical to cardiac output in aortic stenosis. The use of pulmonary artery catheters in the UK is contentious and they are used much less frequently than in the USA [16]. Postoperatively monitoring should continue and it is advised that women with significant cardiac disease should be nursed in a high-dependency unit on the labour ward and cardiac monitoring continued into the puerperium because maternal deaths occur, not uncommonly, 3–5 days postpartum [17]. Postoperative analgesia does not govern the choice of technique for anaesthesia. Patient-controlled analgesia provides satisfactory analgesia after general anaesthesia [18] and by whichever route opioids are administered, respiratory monitoring is required. There have been major advances in diagnostic techniques in the last three decades, improving the ability to monitor patients with cardiac abnormalities serially during pregnancy and continuously during delivery. The principles of successful care depend on maintaining cardiovascular stability with specific cardiovascular monitoring of the functions most at risk. The potential for ensuring the most precise haemodynamic control is better achieved with a general anaesthetic than a spinal technique. Spinal anaesthesia for Caesarean section has a reported incidence of hypotension of up to 83% [19]. Not only is hypotension a hazard in severe aortic stenosis but also the vasopressors and fluid used to treat it are detrimental and, as described above, can commonly lead to circulatory collapse. The woman who had a spinal microcatheter for her Caesarean section was not clinically severely affected by her aortic stenosis and she obviously had abundant cardiovascular reserve. We would recommend general anaesthesia in a more compromised patient.