Vasopressors for the treatment of maternal hypotension following spinal anaesthesia for elective caesarean section: past, present and future.

Abstract

Spinal anaesthesia is the standard technique in many countries when providing anaesthesia for elective caesarean section, as it provides excellent operating conditions and is well tolerated 1. Hypotension remains a common side-effect, and can result in unpleasant symptoms in the mother and harm to the fetus 2. Until fairly recently, ephedrine was the main vasopressor used for the treatment of spinal hypotension. It became the first-line vasopressor following findings from early studies on pregnant ewes that recommended it over metaraminol and other α-adrenoreceptor agonists, as it was associated with less reduction in uterine blood flow 3. An early dose-response study performed by Ngan Kee et al. investigated its use in 80 patients undergoing spinal anaesthesia for elective caesarean section. Patients received either a saline control or 10, 20 or 30 mg of prophylactic intravenous ephedrine. Systolic blood pressure (SBP) after spinal anaesthesia was significantly higher for the 30-mg group compared with other groups. More importantly, however, the proportion of patients found to have an umbilical arterial pH < 7.2 was 11%, 25%, 42% and 22%, in the control, 10-mg, 20-mg, and 30-mg groups, respectively. These findings suggested that although blood pressure control was better with ephedrine than without, there was no improvement in neonatal outcome 4. These concerns prompted the search for a safer vasopressor for use during spinal anaesthesia for caesarean section. Vasopressors with more α-agonist activity had traditionally been considered to be for second-line use only, because of concerns raised from animal studies. Work with human parturients, however, has since shown that both phenylephrine 5 and metaraminol 6 infusions result in improved fetal acid-base status compared with ephedrine. Subsequently, phenylephrine emerged as the vasopressor of choice on the labour ward, as it was more easily available to early investigators and as a result used more widely than metaraminol 7. A randomised, double-blind study performed by Cooper et al. compared three groups of patients undergoing elective caesarean section, receiving infusions of phenylephrine 100 μg.ml−1, ephedrine 3 mg.ml−1, or phenylephrine 50 μg.ml−1 in combination with ephedrine 3 mg.ml−1. They found a lower incidence of fetal acidosis in the groups receiving phenylephrine alone or in combination with ephedrine 5. Subsequently, Ngan Kee and Lee performed a multiple linear regression analysis on 337 consecutive caesarean sections under spinal anaesthesia, investigating different factors that may predict uterine arterial pH and base excess. The use of ephedrine was a significant factor predicting adverse changes in both pH and base excess, and the authors concluded that in order to minimise the risk of fetal acidosis, ephedrine should not be used before delivery, and that α-agonists should be used to minimise spinal hypotension 8. The same group went on to perform a complex study investigating the effect of varying different proportions of vasopressors when used in combination. One hundred and twenty-five parturients undergoing spinal anaesthesia for caesarean section were randomly assigned to receive 100%, 75%, 50%, 25% or 0% phenylephrine with 0%, 25%, 50%, 75% or 100% ephedrine, respectively. Infusions were adjusted to maintain SBP close to baseline. They found that as the proportion of phenylephrine decreased and the proportion of ephedrine increased, haemodynamic control was reduced, and fetal acid-base status was less favourable 9. A recent meta-analysis of vasopressor use during elective caesarean section, by Veeser et al, collated data from 20 trials (n = 1069), finding the relative risk for true fetal acidosis to be 5.29 for ephedrine versus phenylephrine 10. Following such compelling evidence, the use of ephedrine has all but disappeared and phenylephrine has become firmly established as the vasopressor of choice, for both prophylaxis and treatment of spinal hypotension in obstetrics. However, research continues in order to optimise and refine its administration. Areas studied have included: how phenylephrine could best be administered; whether it should be used proactively (prophylactically) or reactively (only when spinal hypotension has occurred); whether continuous infusions are superior to bolus administration; and the appropriate dose or doses required to avoid unwanted side-effects such as reactive hypertension and bradycardia. A meta-analysis looking at the use of prophylactic phenylephrine for caesarean section under spinal anaesthesia 11 concluded that a continuous infusion (proactive treatment) started immediately after initiation of spinal anaesthesia significantly reduced the incidence of spinal hypotension compared with bolus doses given only in response to a fall in SBP (reactive treatment). Prophylactic administration of phenylephrine has been regarded by some as being too aggressive, due to its ability to cause reactive hypertension and associated bradycardia 12. In the meta-analysis described above 11, the risk of reactive hypertension did not differ between prophylactic and reactive regimens, but this result was based on only three studies with a total of 241 patients. The risk of bradycardia was also similar between groups, but again this was based on only small numbers. These results might suggest that there is a paucity of evidence in this area, rather than an absence of an association between phenylephrine and hypertension. However, in the absence of such evidence, prophylactic (proactive) treatment would appear to be preferable, as delaying the start of a prophylactic phenylephrine infusion could limit its efficacy in reducing the incidence of hypotension. Most studies have compared prophylactic phenylephrine infusions with reactive phenylephrine boluses. There are only limited data comparing prophylactic phenylephrine infusions with prophylactic bolus doses. Das Neves et al. compared a prophylactic phenylephrine infusion running at 0.15 μg.kg−1.min−1 with a prophylactic bolus dose of 50 μg phenylephrine given immediately after spinal injection; a third group received the vasopressor only when SBP had dropped. The continuous infusion group had the least incidence of hypotension (18%), nausea (10%), and vomiting (0%) compared with the prophylactic bolus (respectively 33%, 15% and 8%) and therapeutic bolus (respectively 85%, 40% and 13%) groups 13. The higher incidence of hypotension, nausea and vomiting in the prophylactic bolus group may have been because the bolus dose used in the study was small. A study by George et al. 14 found the ED90 of phenylephrine required for the treatment of spinal anaesthesia-induced hypotension to be 150 μg. The ED95 of phenylephrine, found by Tanaka et al. 15, was 159 μg, and the dose to prevent pre-delivery spinal-induced hypotension and nausea at elective caesarean section was 120 μg. Additionally, in the study by das Neves et al., spinal anaesthesia was achieved with only 10 mg bupivacaine 13. These findings support the view that a continuous infusion is superior to bolus administration. A further area to consider is whether phenylephrine, when given by infusion compared with manual bolus administration, can reduce the workload of the attending anaesthetist. Manual bolus doses of a vasopressor to treat hypotension or symptoms of nausea and vomiting certainly can occupy the anaesthetist's attention. A recent study achieved a reduction in anaesthetists’ workload by adhering to an algorithm adjusting the infusion rate of a prophylactic phenylephrine infusion according to changes in blood pressure and heart rate 16. Another yet unresolved issue is the ideal infusion regimen that will control the maternal blood pressure, with minimal maternal side-effects, while avoiding maternal hypertension. Ngan Kee and colleagues conducted their studies infusing phenylephrine at 100 μg.min−1. In one, phenylephrine was infused at 100 μg.min−1 for 3 min following spinal anaesthesia, after which parturients were randomly allocated into two groups. In one, phenylephrine 100 μg.min−1 was infused when the SBP fell below baseline, and this was stopped only if SBP exceeded 120% of baseline. A control group received 100-μg intravenous boluses of phenylephrine after each episode of SBP < 80% of baseline. The infusion group had a reduced incidence of hypotension (23%) compared with the control group (88%). However, hypertension (SBP > 120% of baseline) occurred in 38% of patients in the infusion group compared with only 8% in the control group 17. In the second study by the same group, an infusion of phenylephrine 100 μg.min−1 was started immediately after completion of the intrathecal injection, and was continued for the first 2 min unless SBP exceeded 120% of baseline, in which case it was stopped. After this, the infusion was continued if SBP was less than or equal to baseline, and stopped once it went above baseline. Patients were randomly assigned to two groups depending on the crystalloid infusion received, either a rapid infusion (co-hydration or co-load) group or a minimal maintenance group. Total phenylephrine consumption was lower in the group receiving co-hydration, and hypertension (SBP > 120% of baseline) occurred in almost 50% of patients in both groups 18. Other groups have studied different infusion regimens of phenylephrine ranging from 25 to 100 μg.min−1. Studies by Stewart et al. 19 and Allen et al. 20 both suggested that compared with higher doses, 25–50 μg.min−1 offers the most favourable risk/benefit profile, i.e. the lowest rates of both hypotension and hypertension. Nevertheless, there were two problems with these regimens. First, the need for interventions by the anaesthetist remained high and additional boluses of vasopressor were still necessary in a high proportion of the patients (40%, 20% and 12% in the 25-μg.min−1, 50-μg.min−1 and 100-μg.min−1 groups, respectively) 19. Second, the effect of such rigid haemodynamic control had little beneficial effect on maternal or fetal outcome. The concept of crystalloid co-hydration/co-load and vasopressor use during elective caesarean section under spinal anaesthesia has been a subject of recent interest. Dyer et al. compared crystalloid preload with rapid crystalloid administration after induction of spinal anaesthesia (co-load), finding that coload provided better maternal blood pressure control before delivery 21. A recent review by Ngan Kee further supported its use in the prevention of maternal hypotension after regional anaesthesia 22. Ngan Kee and colleagues further investigated phenylephrine infusions and the optimal blood pressure to which it should be titrated. They randomly allocated parturients to three groups, infusing phenylephrine at 100 μg.min−1 to maintain SBP at 100%, 90% or 80% of baseline. Although patients maintained at 100% of baseline had fewer episodes of hypotension, total doses were higher (1520 μg compared with 1070 μg and 790 μg, respectively). Although higher in the 100% group, umbilical artery pH was always > 7.2. The authors concluded that for optimal management, phenylephrine should be titrated to maintain SBP at near-normal levels. In contrast to their earlier work, high doses of phenylephrine, titrated to 100% baseline, were not associated with maternal hypertension (SBP 120% of baseline) 23. Recent studies have incorporated non-invasive and minimally-invasive cardiac output monitoring during spinal anaesthesia for caesarean section, providing additional insight into the pathophysiology of spinal hypotension in healthy parturients. Both Langesaeter et al. 24 and Dyer et al. 25 used the LidCOplus, the latter group also using transthoracic bioimpedence. Langesaeter et al. used cardiac output monitoring to assess maternal haemodynamic stability in patients receiving high- or low-dose spinal anaesthesia, with or without a concomitant phenylephrine infusion. There was greater haemodynamic stability in patients receiving low-dose spinal anaesthesia combined with a phenylephrine infusion. Dyer et al. compared the effects of phenylephrine and ephedrine boluses on maternal cardiac output. They found that phenylephrine reduced maternal cardiac output compared with ephedrine, and that the fall in cardiac output correlated well with maternal heart rate changes. Stewart et al., using a suprasternal Doppler measurement of cardiac output, found that phenylephrine infusions were associated with a dose-dependent reduction of both heart rate and cardiac output, although no adverse effects on the fetus were seen 21. It would seem therefore that the ideal infusion dose would be one that maintained maternal haemodynamic stability to near baseline, without compromising maternal cardiac output, and current evidence suggests a dose of between 25 and 50 μg.min1. So what else is on the horizon? Closed-loop systems have emerged recently, integrating blood pressure recordings with an infusion pump that responds according to a programmed algorithm, by altering the administration of vasopressor. Sia et al. 26 developed a ‘smart’ system that, when SBP fell below 90% of baseline, administered a 50-μg bolus of phenylephrine; when hypotension occurred with bradycardia, ephedrine was infused. Blood pressure cycling was set at 15 s using a continuous non-invasive technique. No additional vasopressor had to be given by the attending anaesthetist. Ngan Kee et al. compared a computer-based system, infusing 0-100 μg.min1 phenylephrine depending on SBP, with a fixed infusion of 100 μg.min1 phenylephrine that was manually run when SBP fell below baseline, and stopped once SBP (measured at 1-min intervals) exceeded baseline. There were no differences in the incidence of hypotension, hypertension, nausea or vomiting. Only the number of interventions – including starting, stopping, adjusting the computer program or syringe pump, and manual boluses – was different between the groups (median of two in the computer-controlled group and 10 in the manual group) 27. Other vasopressors are being investigated for prophylaxis and treatment of spinal hypotension. In the recent RESPOND study (randomised evaluative study of phenylephrine or noradrenaline for maintenance of blood pressure), 104 healthy women undergoing elective caesarean section under spinal anaesthesia were randomly allocated to receive an infusion of phenylephrine 100 μg.min1 or noradrenaline 5 μg.min1 to maintain maternal blood pressure. The incidence of hypo-/hypertension and nausea/vomiting was low and similar between the groups. Heart rate and cardiac output were greater over time in the noradrenaline group, as were umbilical venous pH and oxygen content, attributed to greater uteroplacental blood flow. The authors raised the idea that noradrenaline (because of its intrinsic β-agonist activity) may be a better obstetric vasopressor than phenylephrine, and recommended further work in this area 28. Use of vasopressors for the treatment and prevention of spinal hypotension has grown in popularity in recent years 29. Ephedrine has largely been superseded by phenylephrine, as the evidence suggests that the former can cause harm to the fetus. We have moved away from a reactive approach (bolus administration once spinal hypotension has occurred) to a more proactive approach, with the advent of phenylephrine infusions. Moreover, we are aiming to limit the amount of vasopressor used to avoid unwanted side-effects such as bradycardia. The ideal vasopressor regimen should allow careful titration to each individual parturient's needs, according to changes in haemodynamic parameters, whilst avoiding excessive demands on the anaesthetist's time. Integrated closed-loop systems with carefully programmed algorithms, as used by Sia et al., seem to come closest to achieving such a goal 25. In the absence of such technology, phenylephrine should be administered as a prophylactic infusion at a dose that prevents maternal hypotension but avoids a significant reduction in maternal heart rate and cardiac output; the literature supports a rate of 25-50 μg.min−1 phenylephrine, titrated to maintain maternal SBP > 80% of baseline, while avoiding maternal hypertension. As for the future: it will be some time before we see routine cardiac output monitoring during elective caesarean section, and it is currently reserved for the management of high-risk parturients, or for research purposes. More work needs to be done investigating the use of noradrenaline infusions for spinal hypotension, and any potential for it to cause maternal or fetal harm, before its use becomes more widespread. Finally, the concept of computer-programmed algorithms could be achievable, if it could be shown to be cost-effective. No external funding and no competing interests declared.