The Effects of Epidural Ropivacaine and Bupivacaine for Cesarean Section on Uteroplacental and Fetal Circulation

The protocol was approved by the Ethics Committee of the University of Oulu and performed in accordance with the principles of the Declaration of Helsinki as amended in Hong Kong, 1989. The patients were studied from October 1991 until June 1992. Twenty-one healthy parturient women with uncomplicated singleton pregnancies at term undergoing elective cesarean section during epidural anesthesia before labor were included in the study. The indications for cesarean section were fetal breech presentation, cephalopelvic disproportion, previous cesarean section, and placenta previa. All participating women were interviewed on the day before the operation; the aim of the study was explained, and informed written consent was obtained.

Each subject received 30 ml 0.3 M sodium citrate orally before transfer to the operating theater. With the patient in the left lateral position on the operating table, the epidural space was identified at the L3-L4 interspace (in one case at L2-L3) with an 18-G needle and the loss-of-resistance technique. A catheter was advanced 3 cm cephalad into the epidural space. After a negative aspiration for blood or cerebrospinal fluid, the catheter was secured, and the patient then resumed a recumbent position with the wedge under the right hip. Local anesthetic was not injected. All patients received oxygen (3 l/min) through nasal prongs, and their heart rhythms were monitored continuously by electrocardiography. Maternal arterial pressures and heart rate were recorded with an automated, noninvasive instrument at first every 5 min and then at 2-min intervals from induction to delivery. Maternal hypotension, defined as a reduction in systolic arterial pressure of more than 30 mmHg relative to the control value (a mean value from three measurements recorded in the operating theater before preloading) or a systolic arterial pressure of less than 100 mmHg, was treated by increasing the rate of infusion, and, if this was ineffective, by intravenous ephedrine.

Fetal heart rate (FHR) was monitored continuously with an external transducer during the induction of the block. FHR tracings were analyzed at the end of the trial by a blinded obstetrician (P.J.). Three characteristics were identified for each tracing: baseline FHR, FHR variability, and periodic excursions of the FHR (accelerations and decelerations). The visual criterion for normal variability in FHR tracings was a heart rate greater than or equal to 5 beats/min. Fetal bradycardia was defined as a heart rate of less than 110 beats/min lasting more than 1 min.

The first Doppler recording was performed immediately after insertion of the epidural catheter before the local anesthetic was given. The equipment and methods used have been described previously. [10]In brief, a color flow mapping system with a 3.75-MHz pulsed-wave Doppler sector transducer (SSH-140A, Toshiba, Tokyo, Japan) was used to obtain blood flow velocity waveforms (the average of three consecutive waveforms) for both maternal uterine arteries (main branch on the placental and on the nonplacental side of the uterus); the placental arcuate artery; and the fetal umbilical, middle cerebral, and renal arteries. Peripheral arterial resistance was expressed in terms of the pulsatility index (PI), calculated as PI = (systolic peak velocity - end-diastolic velocity)/mean velocity during cardiac cycle). The FHR was derived from the umbilical Doppler signal.

An M-mode echocardiographic examination of the fetal heart with the same ultrasound equipment and probe was performed; a four-chamber view of the fetal heart was imaged at the level of the tips of the atrioventricular valves. The end-systolic and end-diastolic inner diameters of the ventricles were recorded from the M-mode image and were used to calculate the fractional and mean circumferential shortenings of the myocardium. [11]The inner end-diastolic diameters were used to determine the size of the fetal heart and the ventricular shortenings to evaluate the fetal heart's functional properties. The aortic and pulmonary valves were then visualized from the five-chamber and short-axis views. The pulsed-Doppler sample volume was placed just above the valve leaflets, and velocity waveforms were measured over three consecutive cardiac cycles and averaged. Recordings with an angle larger than 10 degrees between the transducer beam and the pulmonary trunk and aorta were rejected. All the Doppler recordings were made by the same obstetrician (J.R.) with the subject in the left semilateral position during periods of fetal rest and apnea. The second ultrasound recording lasted an average of 3.9 min (range 3-6 min), and more extensive measurements (the first and third) took an average of 17.6 min (range 13-23 min).

After a rapid infusion of 1 l balanced electrolyte solution, a test dose of 3 ml (15 mg) of the study drug was given epidurally. In the absence of any signs of subarachnoid or intravenous injection, the main dose of 20 ml (100 mg) of the same solution was administered 4 min later, fractionally over an approximately 5-min period. An additional 1 l balanced electrolyte solution was given intravenously while the main dose of the study drug was taking effect. All assessments were related to the end of injection of the main dose (time 0 min). If a sensory block of T6-T4 as assessed by pinprick was not obtained after 20 min, an additional 5 ml (25 mg) was injected. The local anesthetics, 0.5% ropivacaine or 0.5% bupivacaine (Astra Pain Control AB, Sodertalje, Sweden), were supplied in ampules of identical appearance except for the patient number. None of the investigators was aware of which local anesthetic had been injected. The patients were randomized to their treatment group according to a computer-generated randomization.

An ultrasound measurement of the maternal uterine arteries and the fetal umbilical artery (the second ultrasound recording) was carried out 5 min after the main dose of the local anesthetic had been administered. A third recording was performed when sensory analgesia had reached the T6-T4 level by the same methods and with the same targets as in the control measurements.

Sensory block was assessed bilaterally by using analgesia to pinprick (loss of sharp sensation) with a short beveled needle (Sherwood B 400 27G, Astra Pharmaceutical Products). Motor block of the lower limbs was assessed and scored according to a modified Bromage scale: 0 = no paralysis and able to flex the knees and ankles fully; 1 = unable to raise the extended legs and just able to move the knees; 2 = unable to flex the knees and able to flex the ankles; and 3 = unable to flex the ankles. [12]When asymmetric motor blocks were registered, the lower score was chosen. These tests were made at 10 and 20 min and immediately before and after the third ultrasound recording. The occurrence of visceral pain was noted and was defined as a poorly localized, dull, and deep pain or as heaviness, squeezing, tightness, or an unpleasant feeling, as described by the patient. All anesthetic procedures and assessments were performed by the same author (S.A.) according to the same protocol.

Blood samples were obtained from a double-clamped umbilical vein and artery for the determination of blood gases and acid-base values after delivery. Apgar scores were assigned by a pediatrician 1, 5, and 15 min after delivery. Neonatal assessments were performed by examining the neonate's muscle tone, motor behavior in reflex motor responses, and ability to respond appropriately to external events in the environment. Evaluations were made by one of the authors (T.K.-S.) according to the Amiel-Tison-Barrier-Shnider neurologic and adaptive capacity score test (NACS) 15 min and 2 and 24 h after delivery. The NACS is based on 20 criteria, each of which is scored as follows: 0 = absent or grossly abnormal response to testing; 1 = mediocre or slightly abnormal response; or 2 = normal response. These criteria assess five areas: adaptive capacity; passive tone; active tone; primary reflexes; and alertness, crying, and motor activity (general assessments). NACS gives a total score: the maximum is 40, and a score of 35-40 denotes vigor. [13].

Patients were excluded from the study if a bilateral sensory block could not be extended to the T6-T4 level. For statistical analysis of maternal cardiovascular changes, the two-tailed t test for paired data was used. Differences between frequencies in degrees of motor block were analyzed by Fisher's exact test. The null hypothesis of no change in Doppler velocimetry values was tested by using a two-tailed Wilcoxon signed-rank test for paired comparisons in each group and by using a two-tailed Wilcoxon rank-sum test for between-treatment comparisons with logarithmic transformation of the data. Values within groups were compared with the baseline measurement. Associated 90% confidence intervals were computed. Nonparametric point estimates (Hodges-Lehmann) were used. Neurobehavioral data were analyzed by the Mann-Whitney U test. An outcome with a P value less than or equal to 0.05 was considered significant.

Sensory analgesia of no higher than T7 was experienced by one woman in the ropivacaine group, and another patient was therefore recruited to that study group. Because the former patient did not experience pain or discomfort during surgery and had received the treatment according the study protocol it was decided after the trial, but before the data management, to include her data in the statistical analysis. Thus, the series comprised 21 parturient women of whom 11 were given ropivacaine and 10 bupivacaine. The groups were comparable with regard to maternal demographic data (Table 1).

The aim of the treatment was to obtain sensory analgesia to the T6-T4 level. Five patients in the ropivacaine group and three patients in the bupivacaine group were given the maximum dose of 140 mg of the local anesthetic. Visceral pain during surgery was noted in two patients after ropivacaine administration and in four patients after bupivacaine administration. Two patients in the bupivacaine group with inadequate sacral spread of the block requested supplemental intravenous analgesics during surgery, but after delivery. The quality of the surgical anesthesia was judged by the anesthesiologist and the patient as satisfactory in all cases. The mean time to the first request for postoperative analgesia (oxycodone hydrochloride) was 250 min (range 134-459 min) after ropivacaine and 249 min (range 98-780 min) after bupivacaine.

Assessment of the motor block revealed a slower onset after ropivacaine. Six of 11 patients given ropivacaine had a Bromage score of 0 at 10 min after administration of the main dose as compared with 1 of 10 after administration of the main dose of bupivacaine (P < 0.05). At 20 min after administration of the main dose and after the third ultrasound measurement, ropivacaine and bupivacaine were similar with regard to the degree of motor block provided (Table 2).

The mean maternal heart rate was not affected by the administration of either drug, but epidural anesthesia resulted in a significant decrease in mean maternal systolic and diastolic arterial pressures in both groups (P < 0.001). In five patients in the ropivacaine group and in three in the bupivacaine group, the lowest recorded systolic arterial pressure value was less than 100 mmHg. There were no significant differences between the groups with regard to mean maternal arterial pressures (Table 3).

In the bupivacaine group, the uterine artery PI values increased significantly on both the placental and the nonplacental side 5 min after administration of the main dose (P = 0.014, P = 0.002) and when sensory analgesia had reached the T6-T4 level (P = 0.004, P = 0.01) as compared with the baseline measurements. Simultaneously, the PI for the fetal middle cerebral artery decreased significantly (P = 0.02) after bupivacaine administration. There were no significant changes in the Doppler indexes for the fetal renal artery after epidural bupivacaine. After ropivacaine administration the PI for the maternal uterine artery increased significantly (P = 0.01) but only on the nonplacental side and not until sensory analgesia had reached the T6-T4 level. The PIs for the fetal renal and middle cerebral arteries did not change significantly in those who received ropivacaine. No effect on the Doppler indexes obtained from the umbilical artery was observed in either group at any time (Table 4). There were no significant differences relative to baseline values in the fetal myocardial measurements or the systolic peak velocities of the pulmonary trunk and ascending aorta in either group except for a small, just statistically significant decrease in the right ventricular mean circumferential shortening in the bupivacaine group (Table 5). FHR derived from the umbilical Doppler signal remained unchanged in both groups. There were no significant differences between the two treatment groups in any of the ultrasound measurements.

FHR tracings were normal in eight cases after ropivacaine and in seven cases after bupivacaine. Decreased variability was observed in one case and bradycardia in two cases in each group during establishment of the block. One of the bradycardia cases in each group was a minor decrease to 90 beats/min, lasting about 1 min in the ropivacaine-treated patient and 3 min in the bupivacaine-treated patient. Two other cases were major decelerations with a FHR of about 50 beats/min in the ropivacaine-treated patient and of 70-100 beats/min in the bupivacaine-treated patient. Both events lasted about 8 min.

The mean birth weights of the neonates in the ropivacaine and bupivacaine groups were 3,396 g (range 2,440-3,850 g) and 3,801 g (range 2,960-4,740 g), respectively. One infant in the ropivacaine group had a 1-min Apgar score of 8. All other Apgar scores were 9 or 10 at 1, 5, and 15 min in both groups. The mean pH values in the umbilical arterial blood were 7.31 (range 7.26-7.34) after ropivacaine and 7.33 (range 7.28-7.40) after bupivacaine.

The NACS scores, divided into adaptive capacity scores (items 1-5), neurologic scores (items 6-20), and total NACS scores, are presented in Table 6. At 15 min after birth, 6 of 11 ropivacaine-treated and 8 of 10 bupivacaine-treated neonates had total NACS scores equal to or greater than 35; the corresponding values 2 h after birth were 9 of 11 and 9 of 10, respectively. One neonate in the ropivacaine group had a total NACS score of less than 35 at 24 h after birth. There was a statistically significant difference between the groups in neurologic and total NACS scores at 2 h (P < 0.05).

The primary aim of the current study was to examine the influence of 0.5% ropivacaine on the uteroplacental and fetal circulations and on fetal cardiac function and to compare the effects with those of 0.5% bupivacaine by using duplex ultrasound systems. By combining real-time and Doppler techniques and using a color flow system, this method allows noninvasive, detailed, and reproducible evaluation of uterine and fetal hemodynamics. Because absolute volume blood flow measurements are difficult to obtain, most studies have favored qualitative analysis of flow-velocity waveform changes by using various indexes. The indexes most often used are ratios of the maximal frequency shifts in systole and diastole. The increased value of the indexes is related to the increased vascular impedance downstream from the point of measurement, provided that changes in the circulation upstream remain relatively small. [14]The validity of the method has been tested in vitro [15]and with computer models. [16]There is a close correlation between arterial resistance and the PI. [17]The hemodynamic interpretation of the arterial Doppler waveform has been recently reviewed. [18].

It has been demonstrated that ropivacaine and bupivacaine cause different effects in the peripheral vasculature at the injection site, probably as a result of shifts in their dose-response curves for vascular activity. After intradermal injection, 0.75% bupivacaine causes an increase in cutaneous blood flow, whereas the same concentration of ropivacaine decreases it, compared with the effect of saline. [6]In a similar study, the epidural injection of 0.5% bupivacaine caused an increase in epidural blood flow, in contrast to the effect of 0.5% ropivacaine. [8].

The current study demonstrated that 0.5% ropivacaine given epidurally for cesarean section in healthy parturient women had minimal hemodynamic effects on the uteroplacental and fetal circulations, characterized only by a short-term increase in Doppler indexes for the nonplacental uterine artery. Whether the increase in vascular resistance observed here was related to a direct effect on vascular smooth muscle is an open question. Our finding is in agreement with a study in pregnant ewes in which the intravenous infusion of ropivacaine resulted in concentrations in plasma almost twice those expected to occur during routine epidural anesthesia but did not decrease uterine blood flow. [9]Another study in animals has suggested that ropivacaine is slightly less potent than bupivacaine. [19]In the current study, however, ropivacaine and bupivacaine were used in the same concentrations and doses, and the resulting anesthesia was equally effective after both drugs. Despite increased vascular resistance on the nonplacental uterine artery in the ropivacaine group, it is reassuring that the effects, if any, on other parameters recorded in mothers, fetuses, and newborn babies were subtle and transient.

As expected, maternal arterial pressures decreased after epidural ropivacaine and bupivacaine. It has been suggested that maternal hypotension is associated with an increase in the Doppler indexes for the uterine arteries, indicating increased vascular resistance. [20]In that investigation, only a limited fluid preload (10 ml/kg) was used, and accordingly the mean lowest recorded systolic arterial pressure value was 90 mmHg, whereas in the trial presented here the corresponding values were higher, at 103 and 105 mmHg after ropivacaine and bupivacaine, respectively. Furthermore, a systolic arterial pressure reduction to less than 90 mmHg was recorded in 11 of 47 patients in the previous study [20]but in only 1 patient in both groups in the current trial. Therefore, in the current study no real decrease in uteroplacental perfusion as a result of maternal hypotension seems likely.

We observed one case of severe fetal bradycardia after ropivacaine administration. This event was preceded by a maternal nodal rhythm of 44-48 beats/min 8 min after an additional dose of 25 mg ropivacaine. Accordingly, it was assumed that a maternal hemodynamic disturbance was a possible explanation even though no maternal hypotension ensued or signs of systemic toxicity were felt by the patient. Umbilical arterial flow velocity waveforms with continuous end-diastolic flow were recorded continuously during the period of fetal bradycardia. Because the mother's subjective state was satisfactory and because the heart rate of the fetus remained unchanged, although low, we did not proceed with an emergency cesarean section. In such a situation, general anesthesia with its associated further fetal and neonatal drug exposure might induce disastrous sequelae. [21]At delivery, 54 min after the period of fetal bradycardia, the umbilical artery pH value for this neonate was 7.33. Fifteen minutes after birth the baby's total NACS score was 32 but was within normal range at 2 h.

One infant in the ropivacaine group with a low birth weight (2,440 g) scored 30 in the neurobehavioral testing at 15 min and 32 at 24 h after birth. This result may have contributed to the lower NACS scoring in the ropivacaine group because birth weight is of importance in the assessment of the infant's behavior shortly after delivery.* The statistically significant difference in the neurologic scores (the sum of items 6-20) and total NACS scores 2 h after birth is probably not of clinical importance; the actual number of vigorous babies (total score 35 or more) was the same in both groups at that time. All newborns except the newborn with a low birth weight had total NACS scores of 35 or more at 24 h after delivery.

The assessment of motor block characteristics in this study was restricted to characteristics of onset only. The slower onset after ropivacaine is in agreement with animal data. [22]Others have found a slightly shorter duration of the motor block after epidural 0.5% ropivacaine in nonpregnant humans compared with that after bupivacaine of equal concentration. [23]It seems that there is a clinically indistinguishable difference in the motor block characteristics between ropivacaine and bupivacaine in equal concentrations.

An unexpected finding in the current study was that the PIs for both uterine arteries increased after bupivacaine administration. Because the PI is considered a measure of the resistance to blood flow distal to the site of measurement, this finding suggests increased vascular resistance in these vessels. At least two mechanisms may account for this finding. First, bupivacaine has been reported to induce a decrease in the uterine blood flow as a result of a vasoconstriction in the uteroplacental circulation after an intravascular injection in pregnant ewes. [24]After a single epidural injection, arterial plasma bupivacaine concentrations are significantly higher in nonpregnant humans compared with those after fractional administration of a similar dose for the 1st 15 min. [25]Hence, an increase in the PI for the uterine arteries observed in the current study may be a result of bupivacaine concentrations after a bolus injection that were high enough to induce vasoconstriction in these vessels. Second, in dogs arterial pressure is partially maintained by arginine vasopressin release during epidural anesthesia-induced sympathetic block. [26]This hormone increases uterine vascular resistance in gravid ewes. [27]Thus, the more rapid onset of sympathetic block after a single epidural injection may be associated with higher concentrations of vasopressin, which could result in uterine arterial vasoconstriction.

The fetal cardiovascular response to hypoxia is a redistribution of the blood flow and blood oxygen content to favor the fetal heart, brain, and adrenal glands at the expense of the kidney, gut, and musculoskeletal circulation. Arterial resistance diminishes in the cerebral circulation of the hypoxic fetus, a change reflected as a decrease in Doppler indexes. [28]In the current trial, the vascular resistance of the middle cerebral artery decreased after bupivacaine administration. This decrease may be a sign of a fetal blood flow redistribution related to increased vascular resistance in the uterine arteries. However, the PI for the umbilical artery remained stable in both groups.

In this trial, the placental location was determined, and flow velocity waveforms were obtained from both uterine arteries. Accordingly, we were able to show that placental location influenced the blood flow velocity indexes obtained from uterine arteries: the increase in the indexes first appeared and was more marked on the nonplacental side in both groups. In hypertensive pregnant women with unilateral placental location, abnormal uterine artery resistance on the placental side expressed by the Doppler indexes is the best predictor of poor outcome. [29]Further studies are needed to determine whether a selection between placental and nonplacental uterine arteries is of value as an indicator of uteroplacental blood flow compromise during obstetric anesthesia. In any case, this study suggests that when Doppler velocimetry is used to assess the effects of anesthesia medication on uterine circulation it is important first to locate the placenta before uterine artery evaluation.

One mother experienced a major but transient hypotension after bupivacaine. This event was followed by an increase in the Doppler indexes for the uterine arteries. Despite correction of maternal arterial pressure, the uteroplacental resistance did not return to a normal level until several minutes later. This finding agrees with the report of Wright and coworkers. [20]An increase in the PI values in the current study was followed by a fetal bradycardia of 70-100 beats/min for 8 min. Total NACS scores for this baby were initially low, 31 at 15 min, but were within normal range at 2 h after birth. There was another case of fetal bradycardia after bupivacaine, but it was subtle and transient. The incidence of FHR abnormalities is higher in the current small series than that reported previously. [30]Again, the technique of epidural injection may have been a factor accounting for the conflicting results between these two series.

We conclude that in the conditions of the current study, 0.5% ropivacaine proved to be as effective as 0.5% bupivacaine administered epidurally for cesarean section in healthy parturient women and had no detrimental effects on uteroplacental and fetal circulations, apart from a minor increase in the flow velocity waveform indexes for the nonplacental uterine artery. After bupivacaine administration, uterine artery Doppler indexes increased on the nonplacental and on the placental side, indicating increased vascular resistance in these vessels. Neither drug had any detrimental effects on Apgar scores or acid-base values in the umbilical blood flow after delivery. None of the neonates' conditions was markedly depressed according to the neurobehavioral testing. If these results can be ascertained in larger study groups, epidural ropivacaine may prove to be useful in obstetric practice.

The authors thank Sven Sandin, biostatistician, Astra Pain Control AB, for his excellent help with statistics.

* Kangas-Saarela T: The effects of obstetric anaesthesia on neonatal neurobehaviour and spectral EEG. Ph.D. Thesis. Acta Universitas Ouluensis Series D. Medica no 186, 1989, pp 61-62.