Minimal effective weight-based dosing of ondansetron to reduce hypotension in cesarean section under spinal anesthesia: a randomized controlled superiority trial

Spinal anesthesia is now the technique of choice for women undergoing cesarean section. However, hypotension is a common problem after surgery with reported incidences varying from 53 to 85% worldwide [1, 2]. The Bezold–Jarisch reflex, which occurs after spinal anesthesia, induces vasodilatation and decreases venous return which results in bradycardia and hypotension [3]. A systematic review reported that neither intravenous fluid preload nor vasoconstriction given before spinal anesthesia is effective in preventing hypotension [4]. One factor known to influence the Bezold–Jarisch reflex is serotonin. Ondansetron is a serotonin receptor blocker. It decreases serotonin induced by the Bezold–Jarisch reflex by suppressing venodilatation [5]. Two studies found that ondansetron given at a dose of 8 mg before spinal anesthesia could reduce hypotension by 86% in general surgery [6] and by 33% in cesarean section [7]. However, another study showed that the same dose of ondansetron could not significantly reduce hypotension in pregnant women [8]. Fattahi et al. [9] showed that the mean arterial pressure after spinal anesthesia in patients who had a cesarean section was significantly higher in women given ondansetron 0.15 mg/kg compared to the control group. Sahoo et al. [10] reported the lower dose of ondansetron (4 mg) compared to control group could reduce hypotension in parturients in cesarean section. In terms of safety, Pasternak et al. [11] and Einarson et al. [12] showed that ondansetron administered in over 600,000 pregnant women had no significant effect on the newborn.

All of these studies used a fixed dose of ondansetron. To our knowledge, the optimal dose based on the patient’s weight has never been investigated. Using a weight-based dose of ondansetron might maximize its effect and prevent suboptimal dose or overdose in some parturients since they have a larger volume of distribution compared to normal patients. Hence, this study aims to determine the minimal weight-based dose of ondansetron required to reduce hypotension after spinal anesthesia in women undergoing cesarean section.

This was a prospective triple-blinded parallel randomized controlled superiority trial. The study was approved by the Institutional Ethics Committee of The Faculty of Medicine, Prince of Songkla University, Songkhla, Thailand on March 3, 2016 (EC 58382081), Clinicaltrials.​in.​th number TCTR 20160323001 on March 22, 2016.

Two hundred and twenty-eight women were enrolled in the study with details shown in the consort diagram in Fig. 1. Data collection was completed in 215 patients for the intention-to-treat analysis. Patient characteristics and anesthetic data are presented in Table 1. No differences among the groups were observed in demographic data and anesthetic data. The overall incidence of maternal hypotension was 80%. There were no differences in the incidence of hypotension between the three groups either before or after delivery except episodes of hypotension before delivery were significantly higher in group O1 compared to NS (5 vs 2, P = 0.02) (Table 2). The arterial pressures (SBP, MAP, and DBP) are shown in Fig. 2. MAP levels at 4 and 9 min after spinal anesthesia were significantly lower in group O1 compared to the NS group (P = 0.024 and P = 0.034, respectively). The overall heart rates throughout the operations were not different among the three groups (Fig. 3). Ephedrine requirements were similar among the three groups: group NS (6 mg), group O1 (12 mg), and group O2 (6 mg) (P = 0.21). There were also no differences in blood loss among the three groups: group NS (447.9 ± 201.3 mL), group O1 (490.8 ± 241.2 mL), and group O2 (440.3 ± 198.2 mL) (P = 0.32). Table 3 shows the surgical and anesthetic parameters for the three groups. The metoclopramide requirement in the post-hoc analysis in group O2 was significantly lower than in group NS (P = 0.01). Table 4 compares the incidence of maternal and newborn adverse events between the three groups. No significant differences were found although tachypnea was slightly higher among newborns of women who received low dose ondansetron (7.0%, n = 5) compared to the control group (2.8%, n = 2) (P = 0.28). There were no reports of burning or QT prolongation in mothers and no reports of bradycardia or hypothermia in newborns in any group.

In this study, we found that a weight-based dose of prophylactic ondansetron did not influence the incidence of maternal hypotension. The results of this study were consistent with two other studies [8, 13] but were in contrast with others [6, 7, 9, 10]. Two recent systematic reviews [14, 15] showed ondansetron reduced the incidence of hypotension (relative risk ratio [RR] = 0.63, 95% confidence interval [CI]: 0.45–0.88 and RR = 0.67, 95% CI: 0.54–0.83, respectively) and bradycardia (RR = 0.31, 95% CI: 0.19–0.50 and RR = 0.49, 95% CI: 0.28–0.87, respectively) after spinal anesthesia in cesarean section even though a low quality of evidence was reported [15]. However, the minimal effective weight-based dosing of ondansetron in parturients was not determined [9, 10, 14, 15].

In our study, ondansetron had little effect on hemodynamic stability after subarachnoid block in parturients. Factors that impact the level of anesthesia and analgesia given to the patient might also affect the patient’s arterial pressure as well as the way to measure arterial pressure. There are some possible reasons for the difference in the incidence of hypotension compared to other reports. First, we did not monitor invasive arterial blood pressure or cardiac output. Real time arterial pressure monitoring may be better at detecting actual variations of blood pressure than noninvasive blood pressure monitoring although we measured blood pressure every minute during the first 15 min after spinal anesthesia. A noninvasive blood pressure measurement might not be sufficiently sensitive to notice hemodynamic changes to promptly treat hypotension in parturients, especially in preeclamptic patients [16]. Langesaeter et al. [17] and Rosseland et al. [18] showed that continuous invasive arterial pressure monitoring during cesarean section under subarachnoid block provided hemodynamic stability in parturients, especially when vasopressor was given. Second, vasodilatation from subarachnoid block might have a stronger influence on blood pressure than the Bezold–Jarisch reflex in our study. From the results, 0.05 mg of ondansetron caused more episodes of hypotension before delivery compared to the control group (P = 0.02), especially at 4 and 9 min after spinal anesthesia (Fig. 2) which led to failure to prevent the Bezold–Jarisch reflex. Therefore, ondansetron may have little effect on preventing hypotension in these cases since we administered a vasopressor to treat vasodilatation immediately after hypotension occurred causing no differences in the incidence of hypotension among the three groups. We recorded blood pressure after a subarachnoid block was performed every minute for 15 min. If hypotension occurred, ephedrine or norepinephrine was administered instantly until the MAP reached the lower limit of baseline level. The peak onset of ephedrine or norepinephrine is approximately 1–2 min [19] and repeated doses were given as needed. Third, the weight-based dosing of 0.1 mg/kg of ondansetron might not be adequate to maximize the effect to prevent the Bezold–Jarisch reflex. A high dose of ondansetron (12 mg) or 0.15 mg/kg of ondansetron was reported to increase MAP in patients who had a cesarean section under spinal anesthesia compared to the control group [9, 20]. In our study, the mean dose of 3 to 7.9 mg of ondansetron was given which might not be adequate to prevent the Bezold–Jarisch reflex even though some studies found a lower dose of ondansetron (4 or 6 mg) could significantly prevent hypotension in the same setting [10, 20]. However, we were concerned that the dose of 0.15 mg/kg of ondansetron might be harmful to the fetus in terms of possible umbilical arterial vasoconstriction due to ondansetron acting on the 5-HT1B receptor [21]. Fourth, our analysis was based on an intention-to-treat principle which included repeat administration of subarachnoid blocks following failure of the first spinal anesthesia. A repeat dose of subarachnoid block can affect the level of anesthesia and analgesia, although there were no significant differences in the number of blocks given between the three groups (P = 0.87).

The definition of hypotension used in this study was a decrease in the MAP by 30% from the baseline level which was quite similar to many studies [7, 8, 22, 23]. MAP was used instead of SBP to define hypotension because MAP provides a physiologically more appropriate measurement of hypotension than SBP [24]. The overall incidence of hypotension among women undergoing cesarean section following spinal anesthesia in our study (80%) was lower than reported in a previous study done in Songklanagarind Hospital (85%) [2]. One-third of each intervention group received metoclopramide as premedication. Even though a case report showed 10 mg of metoclopramide could cause severe hypotension [25], the baseline blood pressure before spinal anesthesia was not different among the three groups (Fig. 2).

Ondansetron did not cause any adverse events in our study. The parturients experienced neither QT abnormalities nor burning sensation during the injection and the newborns did not experience bradycardia or hypothermia 24 h postoperatively. The incidence of urticaria or arrhythmia was not higher in the two ondansetron groups. Although the incidence of vomiting was lower in the high dose ondansetron (0.1 mg/kg) group, the incidence was not statistically different (P = 0.09). However, postoperative metoclopramide requirement was significantly lower in the high dose ondansetron compared to the normal saline group (P = 0.01). Therefore, the effect of ondansetron to prevent or treat nausea and vomiting was achieved as expected. There were no significant differences in terms of hypotension or tachypnea in newborns between the three groups. The incidence of tachypnea was slightly higher among the newborns of women who received the low dose of ondansetron compared to the control group (7% vs 2.8%). Tachypnea resolved within 2–3 h after oxygen therapy was administered. Although ondansetron can transfer across the placenta to newborn babies [26], adverse events related to ondansetron in newborns have not been reported.

Ondansetron administered at a dose of 0.05 mg/kg or 0.1 mg/kg before spinal anesthesia was not effective in reducing the incidence of hypotension in pregnant women undergoing cesarean section. Higher weight-based dosing of ondansetron should be examined keeping the safety issues in mind.