Use of ephedrine in obstetric patients is associated with depression of fetal acid-base status. The authors hypothesized that the mechanism underlying this is transfer of ephedrine across the placenta and stimulation of metabolism in the fetus.
A total of 104 women having elective Cesarean delivery under spinal anesthesia randomly received infusion of phenylephrine (100 microg/ml) or ephedrine (8 mg/ml) titrated to maintain systolic blood pressure near baseline. At delivery, maternal arterial, umbilical arterial, and umbilical venous blood samples were taken for measurement of blood gases and plasma concentrations of phenylephrine, ephedrine, lactate, glucose, epinephrine, and norepinephrine.
In the ephedrine group, umbilical arterial and umbilical venous pH and base excess were lower, whereas umbilical arterial and umbilical venous plasma concentrations of lactate, glucose, epinephrine, and norepinephrine were greater. Umbilical arterial Pco2 and umbilical venous Po2 were greater in the ephedrine group. Placental transfer was greater for ephedrine (median umbilical venous/maternal arterial plasma concentration ratio 1.13 vs. 0.17). The umbilical arterial/umbilical venous plasma concentration ratio was greater for ephedrine (median 0.83 vs. 0.71).
Ephedrine crosses the placenta to a greater extent and undergoes less early metabolism and/or redistribution in the fetus compared with phenylephrine. The associated increased fetal concentrations of lactate, glucose, and catecholamines support the hypothesis that depression of fetal pH and base excess with ephedrine is related to metabolic effects secondary to stimulation of fetal beta-adrenergic receptors. Despite historical evidence suggesting uteroplacental blood flow may be better maintained with ephedrine, the overall effect of the vasopressors on fetal oxygen supply and demand balance may favor phenylephrine.
REGIONAL anesthesia is normally preferred for Cesarean delivery because it avoids the maternal risks of general anesthesia such as aspiration of gastric contents and difficulty with airway management.1Although it is generally accepted that regional anesthesia confers greater safety for the mother compared with general anesthesia, its effects on neonatal outcome are controversial, particularly for spinal anesthesia. For example, several studies have shown that the risk of fetal acidosis is greater with spinal anesthesia compared with general anesthesia,2and a recent large retrospective study has found that neonatal mortality of very preterm infants born by Cesarean delivery under spinal anesthesia was greater than that of comparable infants delivered under general anesthesia.3The mechanism underlying these observations is uncertain, but recent data suggest that an important contributing factor may be the widespread use of ephedrine to treat and prevent hypotension during regional anesthesia.4,5Historically, ephedrine was recommended as the vasopressor of choice in obstetrics but there is now increasing evidence that ephedrine has the propensity to decrease fetal pH and base excess, especially in comparison with other vasopressors such as phenylephrine4,6and metaraminol.7
The reason why ephedrine is associated with fetal acidosis is unknown. The original recommendations in favor of ephedrine were based on animal and in vitro data that showed that ephedrine has lesser propensity to cause vasoconstriction of the uteroplacental circulation compared with α-adrenergic agonists.8Little regard was given to the possibility that vasopressors may have direct effects on the fetus. Recently, however, it has been postulated that a mechanism explaining the acid-base changes associated with ephedrine is stimulation of fetal metabolism after placental transfer of ephedrine and stimulation of fetal β-adrenergic receptors.5,6However, there are few experimental data available on comparative placental transfer of vasopressors and little information to support this hypothesis. Therefore, we designed this randomized, double-blind study to quantify and compare placental transfer of ephedrine and phenylephrine and the effect of these vasopressors on a number of biochemical markers of metabolism in mother and newborn when used to maintain blood pressure during spinal anesthesia for elective Cesarean delivery.
Materials and Methods
The study was approved by the Joint Chinese University of Hong Kong – New Territories East Cluster Clinical Research Ethics Committee, Shatin, Hong Kong, China, and was registered in the Centre for Clinical Trials Clinical Registry of the Chinese University of Hong Kong (unique trial number CUHK_CCT00079). All patients gave written informed consent. We recruited 104 American Society of Anesthesiologists physical status 1 and 2 women with term singleton pregnancies scheduled for elective Cesarean delivery under spinal anesthesia. Exclusion criteria were hypertension (systolic blood pressure greater than 140 mmHg or diastolic blood pressure greater than 90 mmHg), cardiovascular or cerebrovascular disease, known fetal abnormality, contraindications to spinal anesthesia, and signs of onset of labor.
Patients received oral famotidine and sodium citrate as premedication. On arrival to the operating room, standard noninvasive monitoring was applied, and patients were positioned in the left-tilted supine position for several minutes, during which blood pressure was measured every 1–2 min. Blood pressure measurements were continued until they became consistent (three successive measurements of systolic blood pressure that had a difference of no more than 10%). Baseline systolic blood pressure and heart rate were calculated as the mean of the three recordings.
A 16-gauge IV cannula was then inserted into a forearm vein under local anesthesia and connected by using a wide-bore infusion set to a 1-l bag of warmed lactated Ringer’s solution. No intravenous prehydration was given. Spinal anesthesia was induced in the right lateral position. After skin infiltration with lidocaine, a 25-gauge pencil-point needle was inserted at what was estimated to be the L3-4 or L4-5 vertebral interspace and 2.0 ml of hyperbaric 0.5% bupivacaine (10 mg) and 15 μg of fentanyl were injected intrathecally. The patient was then returned to the left-tilted supine position. Blood pressure was measured at 1-min intervals beginning 1 min after spinal injection. Hemodynamic data were downloaded to a computer at 5-s intervals.
Patients were randomly allocated to have their blood pressure maintained using an infusion of either phenylephrine 100 μg/ml (phenylephrine group) or ephedrine 8 mg/ml (ephedrine group) by drawing of sequentially numbered sealed envelopes that each contained a computer-generated randomization code (Statview for Windows 5.0.1; SAS Institute Inc, Cary, NC). Replacement randomization was used when the codes were generated to ensure equal numbers in each group.9To facilitate double-blinding, the drugs were prepared in identical syringes by one of the investigators who was not involved with subsequent patient management or data collection. The concentrations of vasopressors were chosen on the basis of previously published estimates of comparative potency.10The vasopressors were infused by using a syringe pump (Graseby 3500 Anaesthesia Pump; Graseby Medical Ltd., Watford, Herts, United Kingdom) that was connected via fine-bore tubing to the IV cannula by using a 3-way stopcock. Infusion rates were adjusted to maintain systolic blood pressure near to baseline values by using a previously described regimen.11,12At intrathecal injection, rapid IV fluid infusion (maximum 2 l) was started by fully opening the valve of the infusion set with the fluid bag suspended 1.5 m above the operating table, and the vasopressor was commenced at 60 ml/h. For 2 min, the infusion was continued unless systolic blood pressure was more than 120% of baseline. Subsequently, until terminating the study at uterine incision, systolic blood pressure was measured every 1 min, and the infusion was continued if systolic blood pressure was less than or equal to baseline and stopped if systolic blood pressure was greater than baseline. If there were more than two consecutive episodes of hypotension (defined as systolic blood pressure less than 80% of baseline) a “rescue” IV bolus of 100 μg of phenylephrine was given. The incidence of hypertension, defined as systolic blood pressure greater than 120% of baseline, was recorded.
Five minutes after intrathecal injection, the upper sensory level of anesthesia was recorded by assessing loss of pinprick discrimination, and the surgeon was invited to scrub. Further checks of the block height were made as required before the start of surgery, but these levels were not recorded as part of the study. Surgical times and the incidences of nausea or vomiting were recorded. Supplemental oxygen was not given unless arterial oxyhemoglobin saturation was less than 95%. Bradycardia (heart rate less than 50 beats/min) was treated by stopping the vasopressor and, if accompanied by hypotension, with 0.6 mg of IV atropine.
At the time of delivery, a 5- to 10-ml sample of maternal arterial (MA) blood was taken with a heparinized syringe by radial artery puncture. Immediately after delivery, samples of umbilical arterial (UA) and umbilical venous (UV) blood were taken from a double-clamped segment of cord. With these samples, the following measurements were made: (1) blood gases by using a Rapid Point 400 analyzer (Bayer Diagnostics Mfg [Sudbury] Ltd., Sudbury, United Kingdom); (2) plasma concentrations of lactate and glucose by using the Vitros DT60 II Chemistry System (Ortho-Clinical Diagnostics, Raritan, NJ); (3) plasma concentrations of epinephrine and norepinephrine by using methods described in appendix 1; (4) plasma concentrations of phenylephrine or ephedrine by using methods described in appendix 2.
After delivery, Apgar scores were assessed 1 and 5 min by the attending pediatrician.
No preliminary data for umbilical arterial or venous concentrations of vasopressors were available, so sample size calculation was based on potential differences in UA pH. Assuming SD of 0.0412and anticipated difference of 0.03,4we calculated that a sample size of 38 patients per group would be required to have 90% power with a two-sided α value of 0.05. However, in anticipation from previous experience that obtaining sufficient maternal arterial and umbilical cord blood would likely be difficult in a proportion of cases, the sample size was arbitrarily increased by one-third to give a final sample size of 52 patients per group. Patient characteristics were compared by using Student t test. Other data were compared using the Mann–Whitney U test for unpaired data, the Wilcoxon signed-ranks test for paired data, and the chi-square test or Fishers’ exact test for categorical data. P < 0.05 were considered significant. All analyses were performed using SPSS 220.127.116.11 (SPSS Inc., Chicago, IL).
All 104 patients completed the study. Insufficient blood was obtained from varying numbers of patients in each group for the different assays, as indicated in the tables. Hemodynamic data were excluded from analysis from one patient in each group because severe shivering caused measurement artifacts.
There was no difference between groups in patient characteristics (table 1). Results of analysis of blood gases are shown in table 2. UA and UV pH and base excess were lower in the ephedrine group compared with the phenylephrine group. UA Pco2and UV Po2were greater in the ephedrine group compared with the phenylephrine group.
Plasma concentrations of lactate, glucose, epinephrine, norepinephrine, phenylephrine, and ephedrine are shown in table 3. The plasma concentration of epinephrine was below the lower limit of detection (20 pg/ml) in 16 MA and three UV samples in the ephedrine group and seven MA samples in the phenylephrine group; for the purpose of analysis, each of these results was assigned a value of 19 pg/ml. UA and UV plasma concentrations of lactate, glucose, epinephrine, and norepinephrine were all greater in the ephedrine group compared with the phenylephrine group. MA plasma concentrations of glucose, epinephrine, and norepinephrine were all greater in the ephedrine group compared with the phenylephrine group. UV/MA and UA/UV plasma concentration ratios for phenylephrine and ephedrine are shown in figure 1. For the ephedrine group versus the phenylephrine group, the UV/MA ratio was greater (median 1.13 [interquartile range 1.01–1.23]vs. 0.17 [0.11–0.22], P < 0.001) and the UA/UV ratio was greater (0.83 [0.75–0.91]vs. 0.71 [0.56–0.84], P = 0.001).
Birthweight and Apgar scores were similar between groups. One neonate in the phenylephrine group had an Apgar score of 6 at 1 min; all other scores at 1 min and 5 min were 7 or greater.
Hemodynamic changes, vasopressor use, and the incidence of nausea and vomiting are summarized in table 4. Patients in the ephedrine group received a smaller volume of vasopressor but had higher incidences of hypotension and nausea/vomiting and required more rescue does of phenylephrine. The maximum recorded systolic blood pressure was greater in the ephedrine group, but the minimum recorded heart rate was lower and the incidence of bradycardia was greater in the phenylephrine group. No patient required atropine or supplemental oxygen.
Our study showed that ephedrine crosses the placenta to a greater extent than phenylephrine, as evidenced by considerably greater values for UV/MA plasma concentration ratios in the ephedrine group. Furthermore, the UA/UV plasma concentration ratios were also greater in the ephedrine group, which suggests that ephedrine undergoes early metabolism and/or redistribution in the fetus to a lesser extent compared with phenylephrine. Our results confirmed the results of previous studies that have shown that use of ephedrine is associated with lower values for fetal pH and base excess compared with phenylephrine.4,6In addition, we found that use of ephedrine was associated with greater UA and UV plasma concentrations of lactate, glucose, epinephrine, and norepinephrine and greater UV Pco2compared with phenylephrine. Together, these findings are consistent with the hypothesis that the mechanism underlying the propensity for ephedrine to cause fetal acidosis is transfer of the drug across the placenta and stimulation of metabolic processes in the fetus.
Early studies of vasopressors in obstetrics focused mainly on differences among agents in their effect on uteroplacental blood flow. Animal studies showed that the latter was better maintained with ephedrine compared with α-agonists,8which led to the clinical recommendation that ephedrine should be the vasopressor of choice in obstetric anesthesia because greater uteroplacental blood flow should correspond to greater oxygen supply to the fetus. However, recent clinical data showing that the use of ephedrine is in fact associated with lower values for fetal pH and base excess has led to the reevaluation of older teaching. These data together with the results of our current study suggest that it is also important to consider oxygen demand of the fetus; the most important factor affecting short-term fetal acid-base status is probably the balance between fetal oxygen supply and demand. Although ephedrine may have more favorable effects on fetal oxygen supply compared with phenylephrine, this advantage may be negated by its greater placental transfer and the undesirable effects of this on fetal oxygen demand. When comparing the net effect of vasopressors on fetal oxygen supply and demand balance, we believe that currently available data favor the choice of phenylephrine.
In the current study, UV Po2was lower in the phenylephrine group compared with the ephedrine group, which is similar to findings from our other recent studies.13,14The exact mechanism underlying this observation is uncertain, but a possible explanation is that it may reflect the greater vasoconstrictive effect of phenylephrine on the uteroplacental circulation. Studies in sheep have shown that, although uterine blood flow varies over a wide range, fetal oxygen uptake remains relatively constant, suggesting that the efficiency of oxygen extraction by the fetus is increased when uteroplacental perfusion decreases.15If the same mechanism is present in humans, a reduction of uteroplacental perfusion caused by phenylephrine with an associated increase in fetal oxygen extraction per unit of uteroplacental blood flow would result in decreased uterine venous Po2and thus decreased umbilical venous Po2because the human placental is considered to function as a venous equilibrator. Under normal clinical conditions, any potential adverse effects of phenylephrine on uteroplacental blood flow do not appear to be detrimental because fetal acidosis is not observed with usual clinical doses of phenylephrine. This is consistent with animal studies that have demonstrated that, under normal physiologic conditions, uterine blood flow is in excess of the minimum required to satisfy fetal oxygen demand,16which confers a margin of safety that, to a degree, protects the fetus from fluctuations in uterine blood flow.17However, it is possible that this may not apply in the presence of acute or chronic uteroplacental insufficiency. Therefore, some caution may be prudent when using large doses of phenylephrine in the presence of clinical evidence of fetal compromise, although in a recent comparison of ephedrine and phenylephrine in nonelective Cesarean delivery, we did not find that use of moderate doses of phenylephrine (median dose before delivery, 100 μg; range 0–1,200 μg) to be associated with any evidence of detrimental effects on the fetus.14
The greater placental transfer of ephedrine compared with phenylephrine can be explained by consideration of differences in the molecular structures of these drugs. Both ephedrine and phenylephrine are structurally related derivatives of phenylethylamine. However, unlike phenylephrine, ephedrine lacks hydroxy-substitution of the aromatic ring; in addition, it has an α-methyl substitution of the ethyl sidechain, which phenylephrine lacks. Thus ephedrine can be expected to have greater lipid solubility than phenylephrine, which explains its greater placental transfer. Ephedrine also crosses the blood brain barrier and has central stimulant and appetite suppressant effects.18Of note, our results showed that the median UV/MA plasma concentration ratio for ephedrine was greater than unity, indicating that not only did ephedrine readily cross the placenta; in most patients, UV plasma concentrations were even greater than MA concentrations. This may be the result of ion trapping, which may occur when ephedrine, a basic drug (pKa 9.6), becomes more protonated (ionized) when exposed to the lower pH environment of the fetus, analogous to the effect observed for local anesthetics.19This effect may be accentuated as fetal pH decreases secondary to ephedrine’s metabolic effects.
The metabolic effects of ephedrine on the fetus can be explained by the fact that, in contrast to phenylephrine, ephedrine has significant β adrenoceptor activity.20β stimulation has previously been shown to stimulate metabolism in fetal lambs after isoproterenol administration and in human neonates after administration of terbutaline before Cesarean delivery.21As a result of its indirect action, ephedrine stimulates presynaptic release of norepinephrine, which likely contributed to the higher circulating plasma concentrations of norepinephrine in umbilical blood in the ephedrine group. We also observed maternal effects of ephedrine as evidenced by greater MA plasma concentrations of glucose, epinephrine, and norepinephrine in the ephedrine group.
The ratio of phenylephrine:ephedrine concentrations in the solutions we used assumed a potency ratio of 80:1 as reported by Saravanan et al. 10However, patients in the ephedrine group received a smaller total volume of vasopressor compared with the phenylephrine group, which suggests that the actual potency ratio may be lower. Although patients in the ephedrine group had a greater incidence of hypotension and required more rescue doses of phenylephrine, there was no difference between groups in the minimum recorded systolic blood pressure although maximum recorded systolic blood pressure was greater in the ephedrine group. These varying results illustrate the difficulty in comparing potencies of two drugs that differ in speed of onset and duration of action.
Although the current study and other recent clinical studies have demonstrated that ephedrine is associated with a greater propensity toward fetal acidosis compared with phenylephrine4,6and other α-agonists,7it is uncertain whether this has potential to affect clinical outcome. It should be noted that the majority of published comparative studies, including the current study, have been performed in low-risk elective cases when small differences in anesthetic technique are unlikely to have a major effect on neonatal outcome. It is possible that the metabolic effects of ephedrine on the fetus may be more important in the presence of other factors that may affect the fetal oxygen supply:demand balance, although we found no difference in outcome in our previous study of nonelective Cesarean deliveries.14Of note, a secondary analysis of the large retrospective EPIPAGE study found that neonatal mortality of very preterm infants born by Cesarean delivery under spinal anesthesia was greater than that of comparable infants delivered under general anesthesia (adjusted odds ratio 1.7, 95% confidence interval 1.1 to 2.6) after adjustment for confounding variables, including gestational age and characteristics of the mother, pregnancy, delivery, neonate, and medical management.3The exact mechanism for this finding is uncertain, but severe or sustained hypotension and excessive use of ephedrine were suggested as possible contributing factors. Further work, ideally from prospective studies, is required to confirm these findings.
The authors thank the midwives of the Labour Ward, Prince of Wales Hospital, Shatin, Hong Kong, China for their assistance and cooperation.
Appendix 1: Method Used to Measure Plasma Concentrations of Epinephrine and Norepinephrine
Blood samples were collected and transferred into lithium heparin tubes containing dilute sodium metabisulfite as an antioxidant and were placed in ice. Samples were immediately centrifuged at 4°C, and the plasma was separated and stored at –80°C pending batch analysis. Norepinephrine and epinephrine were measured by using high-performance liquid chromatography with electrochemical detection. The catecholamines were isolated by using alumina adsorption under basic conditions and then reextracted from the alumina using dilute acid solution before their analysis on the high-performance liquid chromatography with electrochemical detection system. The analytes were separated on a reversed-phased column (Ultrasphere intraperitoneal; C18, Beckman Instruments Inc., Fullerton, CA) by using a mobile phase containing methanol-citric acid-EDTA-octane sulfonic acid-water under isocratic condition. Quantitation was then performed by monitoring the drugs by electrochemical detection using a coulometric detector (ESA 5100A; Environmental Science Associates, Bedford, MA). The assay was linear to the lower limit of detection (25 pg/ml for both epinephrine and norepinephrine). There were good linear responses for both epinephrine and norepinephrine, with correlation coefficients better than 0.9980. The lowest limit of detection was at 25 pg/ml at a signal-to-noise ratio of 3. The within-day coefficients of variation for epinephrine and norepinephrine ranged from 5.52% to 10.62% (mean 8.55%) and 2.53% to 10.57% (mean 6.93%), respectively. The between-day coefficients of variation were 6.49% to 10.07% (mean 8.18%) and 7.33% to 13.02% (mean 10.29%), respectively.
Appendix 2: Method Used to Measure Plasma Concentrations of Phenylephrine and Ephedrine
A sensitive method was developed for the analysis of phenylephrine and ephedrine in plasma by using high-performance liquid chromatography-tandem mass spectrometry. Blood samples were collected into heparinized tubes containing sodium metabisulfite as preservative. The tubes were placed in ice, and the blood samples were immediately centrifuged at 4°C. Plasma was collected and stored at –80°C pending batch analysis. A volume of 0.5 ml of plasma was used for both phenylephrine and ephedrine assays. For sample preparation, plasma was transferred into a conical bottom polypropylene tube containing sodium metabisulfite solution and the internal standard, norephedrine HCl (phenylpropanolamine). All analytes were isolated through liquid-liquid extraction by using an organic solvent. The extracts were further purified by back extraction with dilute acid before analysis using high-performance liquid chromatography-tandem mass spectrometry. A set of calibration standards with varying concentrations of the drugs was also prepared and subjected to the same cleanup procedure. The analytes were separated on a reversed phase column (Atlantis dC18; Waters Corporations, Milford, MA) by using acetonitrile and 0.1% formic acid under gradient condition. Quantitation of the drugs was performed by using multiple reaction monitoring (MRM) in positive ionization mode. Phenylephrine and ephedrine were monitored at m/z 168 → 150 and m/z 166 → 148, respectively (API2000; Applied Biosystems, Foster City, CA). The internal standard, norephedrine HCl, was monitored at m/z 152 → 134. Blank plasma showed no interfering peak at the retention time of the analytes studied. The limits of detection for phenylephrine and ephedrine were 0.2 ng/ml and 0.05 ng/ml, respectively, based on a signal-to-noise ratio of 3. Good linear responses were obtained for both phenylephrine (0.2–50 ng/ml) and ephedrine (0.05–500 ng/ml), with correlation coefficient values 0.9960 and 0.9990, respectively. The within-day variation (intraassay) for phenylephrine ranged from 3.90% to 8.90% (mean 6.39%), whereas that of ephedrine ranged from 2.54% to 7.25% (mean 4.93%). The between-day variation (interassay) for phenylephrine ranged from 7.02% to 10.90% (mean 8.88%), whereas that of ephedrine ranged from 4.41% to 8.03% (mean 6.14%). Samples containing ephedrine at a concentration that was outside the calibration curve were diluted with blank plasma and reassayed.
Subsequent to the completion of the initial analysis, it was discovered that norephedrine, the internal standard, was one of the metabolites of ephedrine. Therefore, to ensure accuracy of the results, a secondary analysis was performed. Leftover plasma samples were retested to determine the amounts of norephedrine metabolite present and to quantify its effect on calculated values for ephedrine. To eliminate the area of norephedrine metabolite from the results, aliquots of leftover samples were retested without the addition of external norephedrine. The area of the metabolite and ephedrine was used to obtain the ratio of metabolite over ephedrine (ratio = area of metabolite/area of ephedrine). This ratio was then used to calculate the area of metabolite present in the original assays (area of metabolite in sample = ratio × area of ephedrine). This enabled the area of external norephedrine added as internal standard to be obtained (actual area of norephedrine = apparent area of norephedrine – area of metabolite). By using the new values of norephedrine area, the ephedrine concentrations of samples were then recalculated. This technique was applied to 121 samples for which leftover plasma was available. The results showed that effect of the presence of the norephedrine metabolite in the samples was small, with adjusted values greater than the original values by a mean difference of 2.07% (SD 1.98%). This correction factor was applied to the originally calculated values for which leftover sample was not available (n = 28). All results are given as adjusted values. The values for MA/UV and UA/UV concentration ratios calculated by using adjusted values were virtually identical to values calculated using the original values.