Preterm birth is the leading cause of perinatal morbidity and mortality.1 Both morbidity and mortality rates are inversely related to gestational age at delivery and each day of delay, particularly before 28 weeks of gestation, increases the survival rate by 3%.2 Spontaneous preterm labor with intact membranes is responsible for approximately one-third of all cases of preterm birth.3 Between 24 and 33 weeks of gestation, the benefits of tocolytic therapy generally outweigh the risk of maternal, fetal, or both complications and these agents should be initiated provided no contraindications exist.2
Comparison studies of the effectiveness of different tocolytic drugs showed comparable results. All have demonstrated a relative benefit that consists mainly of prolonging the gestational age for at least 48 hours, which may allow for the administration of corticosteroids and allow maternal transport to a tertiary care center.4 There is no clear first-line tocolytic drug to manage preterm labor. The optimal agent would combine the highest tolerability and the highest proportion of delayed deliveries.2 Most currently available tocolytic agents carry inherent risks of undesirable adverse effects caused by their nonselective pharmacologic actions on maternal or fetal organs. Treatment with β-adrenergic receptor agonists may be accompanied occasionally with severe adverse effects.2 The efficacy of magnesium sulfate is uncertain,5,6 and concerns regarding adverse fetal effects of cyclooxygenase inhibitors limited its use particularly at a gestational age above 30–32 weeks.2 Compared with all of these agents, both nifedipine and atosiban were found to be superior or comparable to β-adrenergic receptor agonists for tocolysis. In addition, both were associated with fewer fetal and maternal adverse effects.7,8 To determine the ideal first-line tocolytic agent, we aimed in this study to compare the tocolytic efficacy and tolerability of nifedipine with that of atosiban among pregnant women with preterm labor.
PATIENTS AND METHODS
This randomized study was held at a single university teaching medical center between January 2008 and December 2011. Women with preterm labor and intact membranes diagnosed between 24 weeks 0 days and 33 weeks 6 days of gestation were included and randomly assigned to either the atosiban group or the nifedipine group. Preterm labor was defined as the presence of four or more uterine contractions, each lasting 30 seconds or more within 30 minutes confirmed by external tocography and accompanied with cervical effacement of at least 50% and cervical dilatation of 0–4 cm among nulliparous women or 1–4 cm among multiparous women. Gestational age was confirmed by either a documented first-trimester ultrasound scan or by a reliable menstrual date confirmed by ultrasonography performed before 20 weeks of gestation. Exclusion criteria included rupture of membranes, vaginal bleeding resulting from placenta previa or placental abruption, fever above 38°C, severe preeclampsia, maternal cardiovascular or liver diseases, systolic blood pressure less than 90 mm Hg, known uterine malformation, intrauterine growth restriction below the fifth percentile, nonreassuring fetal status, antepartum diagnosis of major fetal malformations, multiple gestations other than twins (triplets or greater), and fetal death.
After random allocation to a treatment group, women received atosiban or nifedipine as follows:
1. Atosiban was given as a single loading intravenous dose, 6.75 mg in 0.9% sodium chloride solution, followed by an intravenous infusion of 300 micrograms/min in 0.9% sodium chloride solution for the first 3 hours and then 100 micrograms/min for another 45 hours. The choice of the dose regimen for atosiban was consistent with the recommendations of the product labeling.
2. Nifedipine was given as a loading dose of 20 mg orally followed by another two doses of 20 mg, 20–30 minutes apart as needed. Maintenance was started after 6 hours with 20–40 mg four times a day for a total of 48 hours.
Women were randomly assigned to receive intravenous therapy with atosiban or oral nifedipine for at least 1 hour. Assigned treatment was planned for up to 48 hours. Maintenance therapy after the first 48 hours was not provided. In addition, tocolytic drugs were discontinued when progression to a cervical dilatation of 5 cm or more or rupture of membranes occurred. Other tocolytic agents were not permitted concomitantly. Labor was considered to have progressed and the woman became eligible for rescue tocolysis if either of the following occurred after at least 1 hour of observation during treatment: 1. increase or no change in the frequency of the contractions, or 2. an increase in cervical dilatation of 1 cm or more. If progress was determined after 1 hour or more, but before 48 hours, or if adverse effects of the drug were noted, a crossover of the study drugs was performed and the alternative tocolytic drug (rescue treatment) was initiated. In cases in which both drugs failed before 48 hours of admission, and the gestational age was 28 weeks or less, indomethacin, as a second rescue agent, was then initiated (two 100-mg per rectum tablets, 1 hour apart, followed by oral tablets of 25 mg four times a day for the rest of the 48 hours). Women with successful tocolysis but with a recurrence of preterm labor at any time after the cessation of the study drug were treated again with the same assigned agent, provided that all eligibility criteria were still met. Prophylactic antibiotics for group B streptococcus and corticosteroids were administered according to standard clinical indications.
Uterine contractions were monitored continuously by an external tocodynamometer for 2–4 hours after initiation of the study drug followed by two to three times a day for 30–60 minutes. The primary outcome was to estimate the tocolytic efficacy and tolerability profile that was assessed in terms of the proportion of women who were not delivered and who did not require an alternate tocolytic agent within 48 hours after initiation of the study drug. Secondary outcomes included the proportion of women who did not deliver and did not require an alternate tocolytic agent at 7 days after initiation of therapy. In addition, tocolytic effectiveness was assessed in terms of the total number of women in the intent-to-treat population who had not been delivered at 48 hours and at 7 days after starting treatment, number of preterm deliveries, and the latency period from enrollment until delivery. Safety was assessed by assessing maternal adverse effects that included maternal tachycardia defined as heart rate above 120 beats per minute, hypotension defined as systolic blood pressure less than 90 mm Hg, feeling of palpitations, headache, nausea, vomiting, itching, and rash. Other secondary outcomes examined were gestational age at delivery and neonatal morbidity and mortality related to prematurity that were assessed until either discharge from the hospital or neonatal death.
Randomization of the groups was performed in blocks of 10 using a computer randomization sequence generation program and the randomization results were kept in the labor and delivery ward in a closed study box. The sequence was concealed until intervention was assigned, ie, just before administrating the tocolytic drug. Because the study drugs were administered by different roots, blinding of participants or care providers was not performed. All authors took part in enrolling and assigning women to interventions.
The study was approved by the local institutional review board, Emek Medical Center, and the Israeli ministry of health. Each woman signed an informed consent.
An intent-to-treat analysis was performed. To assess the difference in the two treatment groups’ demographic and clinical data, the Wilcoxon rank-sum test was used for continuous data and the χ2 or Fisher’s exact tests where appropriate were used for the categorical data. Adjusted odds ratio and 95% confidence interval were calculated for the tocolytic efficacy and tolerability and for the tocolytic effectiveness at 48 hours and 7 days by multiple logistic regression analysis. The same calculations were performed for other secondary outcomes. Adjustment was made for twins, previous preterm delivery, progesterone treatment, closed cervix at presentation, and additional tocolytics. Kaplan-Meier survival analysis was also performed for comparing gestational age at birth using the intent-to-treat protocol. Breslow’s test (Gehan's generalized Wilcoxon test) was used to compare the survival time of the two treatments. This test gives greater weight for differences in treatment at the beginning of the study than at the end of the study time.
Sample size was based on the results of a previous multicenter study that our department took part in8 and which showed that the tocolytic efficacy and tolerability of atosiban is 71.4% at 48 hours. Accordingly 70 women per group were sufficient to show a difference of 25% in the tocolytic efficacy and tolerability of atosiban as compared with nifedipine (ie, 71.4% to 46.4%) with a level of significance of 95% (α=0.05) and a power of 80% (β=0.2).
One hundred forty-nine women were randomized during the study period. Three women were excluded as a result of cervical progression beyond 4 cm before the study drug was initiated and one woman allocated to atosiban withdrew her consent before therapy. All other women were included and analyzed, 75 women in the nifedipine group and 70 in the atosiban group (Fig. 1). Baseline demographic and obstetric characteristics were comparable between the groups (Table 1). None of the women had a cervical cerclage.
The tocolytic efficacy and tolerability profile at 48 hours was significantly higher among women allocated to atosiban as compared with the nifedipine (Table 2). Of all women allocated to atosiban, 48 (68.6%, 95% confidence interval [CI] 57.0–78.6) did not deliver and did not require an alternate tocolytic agent at 48 hours as compared with 39 (52.0%, 95% CI 40.7–63.1) among the nifedipine group (P=.03). This difference was not significant at 7 days from enrollment. Tocolytic effectiveness that was assessed in term of the proportions of women in the intent-to-treat population who had not been delivered at 48 hours after treatment regardless of whether a crossover was made or not did not differ between the groups irrespective of gestational age at enrollment or the number of fetuses (Table 2). Nevertheless, the proportion of women who remained undelivered at 7 days had a singleton and enrolled at 28 weeks of gestation or more was significantly higher among the nifedipine compared with the atosiban group (P=.02) (Table 2). Of all women who had a crossover to the alternate agent within 48 hours from enrollment, only one was the result of adverse effects within the nifedipine group. Two women in each group required the use of a third line, ie, indomethacin. Nine women (12.9%, 95% CI 6.5–22.3) in the atosiban group and 11 (14.7%, 95% CI 8.0–24.1) in the nifedipine group received one or more additional treatments, 48 hours or more after enrollment (P=.9). There was no difference between the groups with any particular maternal morbidity investigated (Table 3); however, the number of women with any adverse effect was significantly higher among the nifedipine (22.7%, 95% CI 14.3–33.2) compared with the atosiban group (7.1%, 95% CI 2.7–15.1) (adjusted odds ratio [OR] 4.18, 95% CI 1.38–12.68, P=.01). Women with any cardiovascular adverse effect (hypotension, tachycardia, or palpitation) was also significantly higher among the nifedipine (17.3%, 95% CI 14.3–33.2) compared with the atosiban group (4.3%, 95% CI 2.7–15.1) (adjusted OR 5.14, 95% CI 1.23–21.40, P=.03).
Gestational age at delivery was significantly higher among the nifedipine group compared with the atosiban group, regardless of whether labor was initiated spontaneously or not (Table 4). The number of women delivered preterm, ie, less than 37 weeks of gestation, was significantly higher among the atosiban group (66.2%, 95% CI 54.4–76.6) as compared with the nifedipine group (43.2%, 95% CI 32.0–54.7, P=.007). The incidence of deliveries at less than 34 weeks of gestation did not differ significantly between the groups (Table 4).
Among the nifedipine group, 40 (53.3%, 95% CI 42.0–64.4) women used the assigned drug only (delivery was delayed for more than 48 hours without a rescue or dilatation progressed to more than 5 cm within 1 hour from enrollment) compared with 54 (77.1%, 95% CI 66.2–85.8) women among the atosiban group (OR 2.9, 95% CI 1.4–6.1, P=.003). Median gestational age at delivery within these pure groups was also higher among the nifedipine group (37.6 weeks, range 28.9–41.0 weeks) compared with the atosiban group (35.6 weeks, range 24.7–40.3 weeks, P=.02).
Kaplan-Meier survival analysis of the gestational age at birth showed that there was a statistically significant difference between the two treatment groups (Breslow P=.04). Median gestational age for women allocated to the nifedipine group was 37 weeks (95% CI 36.2–37.4) and 35 weeks (95% CI 34.1–35.6) for women allocated to the atosiban group (Fig. 2). Similar results were found when stratifying for gestational age at enrollment (P=.03).
The number of neonates admitted to the neonatal intensive care unit (NICU) and length of stay after delivery were significantly higher among the atosiban group as compared with the nifedipine group (Table 4). Overall, neonatal morbidity related to prematurity was comparable between the groups, although the incidence of respiratory distress syndrome and mechanical ventilation among singletons was more than double within the atosiban group as compared with the nifedipine group; however, this difference was not significant. There were no neonatal deaths in either study group (Table 5).
The present study was undertaken to compare the tocolytic efficacy and tolerability profile of nifedipine as compared with atosiban among pregnant women admitting with signs of preterm labor between 24 and 34 weeks of gestation. Atosiban was found to be superior to nifedipine in terms of the proportion of women who did not give birth and who did not require an alternate tocolytic agent within 48 hours from the initiation of therapy. In addition, atosiban was better tolerated by women compared with nifedipine. Both drugs were comparable in terms of the total number of women, in the intent-to-treat analysis, who did not deliver at 48 hours after enrollment. However, the proportion of women who had a singleton and enrolled at 28 weeks of gestation or more and who had remained undelivered at 7 days was significantly higher among the nifedipine compared with the atosiban group. Additionally, women assigned to nifedipine had a lower rate of preterm deliveries and delivered 1 week more than women assigned to atosiban. As a result, probably, more neonates in the atosiban group were hospitalized at the NICU and for a longer time. Furthermore, the incidence of respiratory distress syndrome and mechanical ventilation among singletons was more than two times greater within the atosiban group as compared with the nifedipine group, although this difference was not significant.
Preterm birth is responsible for approximately 75% of all neonatal deaths and 50% of childhood neurologic morbidities.9,10 It is also associated with both high immediate and long-term costs after discharge from the hospital.11 Postponing delivery for 48 hours to administer corticosteroids or to gain critical time to allow the transfer of women to a center with NICU facilities or both is the crucial goal to achieve with tocolytic drugs in cases of a threatening preterm delivery. Several agents have been used to inhibit uterine contractility, but it remains unclear what is the first-line tocolytic agent. Betamimetics reduce the rate of preterm delivery within 48 hours; however, maternal adverse effects are considerable.7 Magnesium sulfate has not been proven to be an effective tocolytic agent at delaying birth, and its use had been reported to be associated with an increased incidence of adverse effects and risk of neonatal mortality.12 Concerns regarding adverse effects of cyclooxygenase inhibitors on the fetal kidneys, ductus arteriosus, increased risk of neonatal intraventricular hemorrhage, and necrotizing enterocolitis have limited its use, especially at a gestational age above 30–32 weeks.2 Nifedipine and atosiban were found to be comparable to betamimetics and both were associated with significantly fewer maternal adverse effects as compared with betamimetics.7,8
Atosiban is currently licensed in Europe but not in the United States. As a result of its placebo-like maternal-fetal adverse effect profile, the Royal College of Obstetricians and Gynecologists suggested considering atosiban as a first-line tocolytic agent.2,8,13,14 On the other hand, several systematic reviews and meta-analysis confirmed that in addition to its comparable effect to betamimetics in delaying delivery for more than 48 hours, nifedipine reduced the incidence of preterm delivery, and more notably reduced neonatal morbidity, including respiratory distress syndrome and NICU admissions.7,15 Although nifedipine is not licensed as a tocolytic agent, the results of these meta-analyses and others brought some authors to suggest that nifedipine appears to meet several characteristics of an ideal tocolytic agent and may be considered a first-line tocolytic agent.15,16
In view of this, the question of which agent should be the first line, atosiban or nifedipine, is still a subject of controversy in many articles. A meta-analysis with an indirect comparison of nifedipine and atosiban showed that the former was associated with a nonsignificant increase in the number of women whose delivery was delayed for at least 48 hours and a significant reduction in respiratory distress syndrome compared with atosiban. Because biases may go along with results of a meta-analysis with an indirect comparison, the authors strengthen the need for better evidence from randomized trials directly comparing nifedipine and atosiban.17 Following the publication of this meta-analysis, two small, underpowered studies that compared the efficacy and safety of atosiban and nifedipine in preventing or delaying preterm labor were published. According to the results of both studies, the efficacy of both medications was the same, but adverse effects of nifedipine were significantly more than atosiban.18,19
In this study, atosiban as compared with nifedipine was found to be superior in terms of the proportion of women who did not deliver and who did not require an alternate agent within 48 hours after initiation of therapy. This effect was pronounced, particularly among women admitted at a gestational age of 28 weeks or more. Similar results regarding the inferiority of atosiban at less than 28 weeks of gestation were reported by others.1,19 This observation may be attributed to the fact that lower concentrations of oxytocin receptors are present at an earlier gestational age.20 In addition, the onset of preterm uterine contractions, particularly the very preterm, may involve other pathways that do not bring into play oxytocin.21 On the other hand, nifedipine as compared with atosiban was found to be associated with a higher mean gestational age at delivery, a lower incidence of preterm deliveries, and, probably accordingly, lead to a lower incidence of respiratory distress syndrome, NICU admissions, and a shorter length of stay. This effect was valid among women assigned to nifedipine regardless of whether they received nifedipine only or needed a rescue agent. Although both drugs were discontinued at 48 hours from enrollment, these repeated observations regarding the superiority of nifedipine may indicate that it probably has an additional long-acting effect that may involve other pathways in the pathophysiologic process of preterm birth beyond that achieved by impeding acute uterine contractions.
The limitations of the current study are worth mentioning. First, although we conducted a priori sample size calculation to ensure adequate power to examine the efficacy and tolerability of the tocolytic drugs, we may be underpowered to detect small differences in terms of secondary outcomes. Second, we did not use a third agent as a first rescue drug; rather, we chose to make a crossover between the study drugs. The decision was based on medical and ethical issues, because both study drugs are considered to have the greatest maternal and fetal safety profile compared with other medications.
The choice of the first-line tocolytic agent in terms of safety profile, efficacy, and costs is a topic of debate. An ideal tocolytic agent is supposed to delay delivery with minimal maternal and fetal adverse effects at low costs. None of the tocolytic medications available fulfills all these criteria. According to the current study, atosiban has fewer failures within 48 hours and probably a better maternal safety profile, both of which favor its use as a first line, particularly in cases in which maternal transport to a tertiary center is being planned or among women with cardiovascular disorders. On the other hand, the overall low incidence of maternal adverse effects associated with the use of nifedipine, the oral route of administration, low costs compared with atosiban, better effectiveness, and a possible efficacy in reducing neonatal morbidity favor its use as a first line, ie, the first agent to start with, among the remaining cases.
1. Romero R, Sibai BM, Sanchez-Ramos L, Valenzuela GJ, Veille JC, Tabor B, et al.. An oxytocin receptor antagonist (atosiban) in the treatment of preterm labor: a randomized, double-blind, placebo-controlled trial with tocolytic rescue. Am J Obstet Gynecol 2000;182:1173–83.
2. Meloni A, Melis M, Alba E, Deiana S, Atzei A, Paoletti AM, et al.. Medical therapy in the management of preterm birth. J Matern Fetal Neonatal Med 2009;22(suppl):72–6.
3. Arias F, Tomich P. Etiology and outcome of low birth weight and preterm infants. Obstet Gynecol 1982;60:277–81.
4. Management of preterm labor. Practice Bulletin No. 127. American College of Obstetricians and Gynecologists. Obstet Gynecol 2012;119:1308–17.
5. Grimes DA, Nanda K. Magnesium sulfate tocolysis: time to quit. Obstet Gynecol 2006;108:986–9.
6. Mittendorf R, Pryde PG. A review of the role for magnesium sulphate in preterm labour. BJOG 2005;112(suppl):84–8.
7. Conde-Agudelo A, Romero R, Kusanovic JP. Nifedipine in the management of preterm labor: a systematic review and metaanalysis. Am J Obstet Gynecol 2011;204:134.e1–20.
8. Moutquin JM, Sherman D, Cohen H, Mohide PT, Hochner-Celnikier D, Fejgin M, et al.. Double-blind, randomized, controlled trial of atosiban and ritodrine in the treatment of preterm labor: a multicenter effectiveness and safety study. Am J Obstet Gynecol 2000;182:1191–9.
9. Slattery MM, Morrison JJ. Preterm delivery. Lancet 2002;360:1489–97.
10. Hack M, Dandruff AA. Outcomes of children of extremely low birthweight and gestational age in the 1990’s. Early Hum Dev 1999;53:193–218.
11. Petrou S, Mehta Z, Hockley C, Cook-Mozaffari P, Henderson J, Goldacre M. The impact of preterm birth on hospital inpatient admissions and costs during the first 5 years of life. Pediatrics 2003;112:1290–7.
12. Crowther CA, Hiller JE, Doyle LW. Magnesium sulphate for preventing preterm birth in threatened preterm labour. The Cochrane Database of Systematic Reviews 2002, Issue 4. Art. No.: CD001060. DOI: 10.1002/14651858.CD001060.
13. Usta IM, Khalil A, Nassar AH. Oxytocin antagonists for the management of preterm birth: a review. Am J Perinatol 2011;28:449–60.
14. Petraglia F, Visser GH. Prevention and management of preterm labour. J Matern Fetal Neonatal Med 2009;22(suppl):24–30.
15. King JF, Flenady V, Papatsonis D, Dekker G, Carbonne B. Calcium channel blockers for inhibiting preterm labour; a systematic review of the evidence and a protocol for administration of nifedipine. Aust N Z J Obstet Gynaecol 2003;43:192–8.
16. Tsatsaris V, Papatsonis D, Goffinet F, Dekker G, Carbonne B. Tocolysis with nifedipine or beta-adrenergic agonists: a meta-analysis. Obstet Gynecol 2001;97:840–7.
17. Coomarasamy A, Knox EM, Gee H, Song F, Khan KS. Effectiveness of nifedipine versus atosiban for tocolysis in preterm labour: a meta-analysis with an indirect comparison of randomised trials. BJOG 2003;110:1045–9.
18. Kashanian M, Akbarian AR, Soltanzadeh M. Atosiban and nifedipin for the treatment of preterm labor. Int J Gynaecol Obstet 2005;91:10–4.
19. Al-Omari WR, Al-Shammaa HB, Al-Tikriti EM, Ahmed KW. Atosiban and nifedipine in acute tocolysis: a comparative study. Eur J Obstet Gynecol Reprod Biol 2006;128:129–34.
20. Fuchs AR, Fuchs F, Husslein P, Soloff MS. Oxytocin receptors in the human uterus during pregnancy and parturition. Am J Obstet Gynecol 1984;150:734–41.
© 2012 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
21. Cole RM, Lamont RF. Current perspectives on drug treatment for preterm labour. J Obstet Gynaecol 1998;18:309–14.