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

Pittard and Vucevic (Anaesthesia 1998; 53: 169–191), describing the use of a subarachnoid microcatheter technique for Caesarean section in a patient with aortic stenosis, correctly assert that the anaesthetic management of such patients for noncardiac surgery is a complex and contentious matter. Most women seen in UK maternity units with aortic stenosis will have congenital rather than acquired disease and many will be, or will have previously been, under the care of a cardiologist. Congenital aortic stenosis has been considered a relative contraindication to pregnancy because of the high maternal mortality (17%) previously reported [1], although this figure has been disputed [2]. The underlying pathophysiology is well known; the stenotic valve leads to left ventricular hypertrophy and eventual myocardial ischaemia and there is a relatively fixed cardiac output. Haemodynamic stability is essential in the safe anaesthetic management of patients with aortic stenosis. Tachycardia decreases the time available for coronary filling; acute reduction in afterload can also severely compromise coronary perfusion. The need to maintain afterload has been extrapolated to suggest that regional analgesia and anaesthesia are contraindicated in pathological states producing a fixed cardiac output. Authoritative sources state ‘high subarachnoid or extradural blockade is contraindicated in patients with cardiovascular disease. By and large … SAB (subarachnoid blockade) and extradural blocks should be avoided in patients with severe cardiac disease’ [3]. The author qualifies his statement with the phrase ‘for anaesthetists with little experience in regional anaesthesia’ and this is surely the relevant point. The inexperienced anaesthetist should not be administering anaesthesia, either regional or general, for Caesarean section, particularly to the woman with aortic stenosis. The obstetric anaesthetist must be completely familiar with regional and general anaesthesia for uncomplicated Caesarean section before undertaking high-risk cases. The Confidential Enquiries into Maternal Deaths in the UK has identified women with coexisting medical disease as a high-risk group in danger of substandard care. It is possible to provide safe regional anaesthesia for Caesarean section in women with aortic stenosis, but certain conditions must be met. Firstly, it is essential to identify these women as early as possible in the antenatal period. There is no place for assessing the woman with aortic stenosis a few minutes before she is due to be delivered. The key to successful management is regular assessment of cardiac status during pregnancy and aggressive management of any deterioration. Secondly, it is important to make clear written and dated management plans for anaesthesia and delivery in the patient's records. These plans should include provision for emergency delivery and should be amended as necessary during the course of the pregnancy. The management plan must stipulate the involvement of senior medical staff, as lack of such involvement has been clearly identified as a factor in substandard care of women dying from childbirth [4]. Regular revision of management plans is important, since deterioration in cardiac status during pregnancy increases maternal risk. Lao and colleagues [5] have published a retrospective review of 25 pregnancies in 13 women with aortic stenosis delivered in Toronto between 1976 and 1992. Although the criticisms of retrospective analysis apply, it is interesting to note that of the 20 pregnancies that continued to term, six ended with Caesarean delivery, all using epidural anaesthesia. Epidural analgesia was also offered for all vaginal deliveries. Lao's group identify coexisting aortic coarctation, symptomatic aortic stenosis at onset of pregnancy and deterioration in cardiac status during pregnancy as the most important risk factors for women with aortic stenosis. They suggest that women who have symptomatic and echocardiographic evidence of severe disease in early pregnancy and those whose cardiac function deteriorates during the first trimester should be offered therapeutic termination of pregnancy. The remaining women should be closely monitored by both obstetricians and cardiologists (and, I would suggest, obstetric anaesthetists) throughout the pregnancy. They advise that echocardiography should be performed in the first trimester of pregnancy and repeated at least once during the second and third trimesters. Echocardiographic estimation of aortic valve area has been shown to be a better index of severity of disease during pregnancy than pressure gradient estimation, since high blood flow in pregnancy can exaggerate estimations of pressure gradient [6]. The increase in blood volume which occurs during pregnancy can precipitate left ventricular failure and pulmonary oedema. Any suspicion of left ventricular failure should be aggressively managed antenatally. Pittard and Vucevic's patient satisfies Lao's criteria for relatively low risk in early pregnancy. She was asymptomatic when seen by her cardiologist before conception. Her electrocardiograph was normal and a repeat echocardiograph at 16 weeks' gestation demonstrated only a slight increase in pressure gradient across the aortic valve. The valve area was apparently not estimated. Ideally she should have had a further echocardiograph performed during the third trimester; however, she remained asymptomatic throughout pregnancy, specifically having no evidence of cardiac failure by 36 weeks' gestation. Lao reported the use of pulse oximetry and invasive arterial and central venous pressure monitoring in all his patients. Pittard and Vucevic also established this level of continuous monitoring, plus continuous electrocardiography. Assuming that antenatal management has been optimal and that the additional risk factors mentioned have been excluded, why should regional anaesthesia be avoided for Caesarean section? Didactic protocols prescribing general anaesthesia for patients with aortic stenosis presuppose the inevitability of uncontrolled decrease in cardiac output if regional anesthesia is performed. Pittard and Vucevic outline the potential problems of single-shot spinal anaesthesia. Rapid onset of sympathetic blockade accompanied by peripheral vasodilatation may cause a sudden decrease in cardiac output, with potentially fatal results. Aggressive infusion of fluids intravenously to treat this situation may compound problems for the anaesthetist by precipitating left ventricular failure. Similar problems may be encountered by the unprepared anaesthetist administering general anaesthesia for Caesarean section. A recently published major obstetric anaesthetic textbook recommends a high-dose opioid technique (fentanyl 15 μg.kg−1) plus etomidate for induction of general anesthesia in the pregnant cardiac patient, with supplementary etomidate but no volatile agent for maintenance [7]. Use of a ‘standard’ obstetric general anaesthetic — a rapid sequence induction with an untitrated intravenous bolus of thiopentone followed by intermittent positive pressure ventilation with nitrous oxide and a volatile agent in oxygen — may result in a sudden decrease in cardiac output and the same disastrous consequences that may ensue after a ‘standard’ obstetric spinal anaesthetic. The solution, whether using regional or general anaesthesia, is to avoid standard techniques and instead to tailor the anaesthetic to the patient. Single-shot spinal is not a suitable regional technique in this situation, but the regional armamentarium of the competent obstetric anaesthetist also includes epidural and combined spinal/epidural anaesthesia. Pittard and Vucevic suggest that continuous spinal anaesthesia should also be considered. Rapid onset of sympathetic blockade can be avoided by using an incremental technique to induce regional anaesthesia (either epidural or spinal). Continuous monitoring of arterial and right atrial pressures permits rapid correction of hypotension. Volume loading can be minimised by administration of vasopressors. Pittard and Vucevic used ephedrine; a study of left ventricular filling dynamics in nonpregnant patients with aortic stenosis demonstrated that administration of a bolus of phenylephrine improved left ventricular filling [8]. Other studies have suggested that a volume preload is not necessary to prevent spinal-induced hypotension at Caesarean section if vasopressors are used [9, 10]. Use of the wedged supine position, or of lateral tilt to ensure displacement of the uterus off the aorta and inferior vena cava, is mandatory during anaesthesia for Caesarean section. It is important to remember that the tilted or wedged position is a compromise between the full lateral position which prevents aortocaval compression and the supine position that the obstetricians would prefer to facilitate surgery. Maintaining the full lateral position until the obstetrician is ready to perform skin incision reduces the risk of haemodynamic instability due to aortocaval compression. Pittard and Vucevic suggest that severity of stenosis is not a factor influencing the decision to use regional or general anaesthesia; Lao considers that deterioration in functional cardiac status during pregnancy is a significant risk factor for poorer outcome, although he does not suggest that regional anesthesia be avoided in such women. The various case reports in the literature tend to be of women with symptomatic or mixed valve disease, often requiring cardiac intervention [11, 12] and it may be that haemodynamic stability is more readily maintained with general anaesthesia in these women, although there is only anecdotal evidence to support this. Ultimately there is, of course, no randomised controlled data comparing carefully managed regional anaesthesia with ‘cardiac’ general anaesthesia for Caesarean section in women with aortic stenosis. All the arguments for and against both techniques are based on anecdotal case reports and assessment of theoretical risks. Multidisciplinary antenatal care of these women is important and should involve senior obstetricians, anaesthetists and cardiologists and regular assessment of cardiac function. Anaesthetic plans for delivery should include provision for the use of invasive monitoring in the peri- and postoperative period and for high-dependency care postoperatively. Both general and regional anaesthesia have significant risks, but incremental induction of either epidural or spinal anaesthesia should be considered a reasonable alternative to general anaesthesia for Caesarean section in the woman with aortic stenosis.