An oxytocin receptor antagonist (atosiban) in the ... -

An oxytocin receptor antagonist (atosiban) in the ... -

An oxytocin receptor antagonist (atosiban) in the treatment ofpreterm labor: A randomized, double-blind, placebo-controlledtrial with tocolytic rescueRoberto Romero, MD, Baha M. Sibai, MD, Luis Sanchez-Ramos, MD, Guillermo J. Valenzuela, MD,Jean-Claude Veille, MD, Bannie Tabor, MD, Kenneth G. Perry, MD, Michael Varner, MD, T. MurphyGoodwin, MD, Rosanne Lane, MAS, Judith Smith, PhD, Gary Shangold, MD, and George W. Creasy, MDDetroit, Michigan, Memphis, Tennessee, Jacksonville, Florida, San Bernardino and Los Angeles, California, Winston-Salem,North Carolina, Fort Worth, Texas, Jackson, Mississippi, Salt Lake City, Utah, and Raritan, New JerseyOBJECTIVES: This study was designed to evaluate the efficacy and safety of the oxytocin receptor antagonistatosiban in the treatment of preterm labor.STUDY DESIGN: A multicenter, double-blind, placebo-controlled trial with tocolytic rescue was designed.Five hundred thirty-one patients were randomized to receive, and 501 received, either intravenous atosiban(n = 246) or placebo (n = 255), followed by subcutaneous maintenance with the assigned agent. Standard tocolyticsas rescue tocolysis were permitted after 1 hour of either placebo or atosiban if preterm labor continued.The primary end point was the time from the start of study drug to delivery or therapeutic failure.Secondary end points were the proportion of patients who remained undelivered and did not receive an alternatetocolytic at 24 hours, 48 hours, and 7 days.RESULTS: No significant difference was found in the time from start of treatment to delivery or therapeuticfailure between atosiban and placebo (median, 25.6 days vs 21.0 days, respectively; P = .6). The percentagesof patients remaining undelivered and not requiring an alternate tocolytic at 24 hours, 48 hours, and7 days were significantly higher in the atosiban group than in the control group (all P ≤ .008). A significanttreatment–by–gestational age interaction existed for the 48-hour and 7-day end points. Atosiban was consistentlysuperior to placebo at a gestational age of ≥28 weeks. Fourteen atosiban-treated patients and 5placebo-treated patients were randomized at

1174 Romero et al May 2000Am J Obstet Gynecolfetal side effects. 7-10 Although other compounds havebeen used for tocolysis, questions remain as to their efficacyand safety, 11-14 Oxytocin is a potent stimulant of myometrialcontractility. A role for this hormone and its receptorin human parturition has been postulated. 15-19Atosiban (Antocin, R.W. Johnson Pharmaceutical ResearchInstitute, Raritan, NJ) is a selective oxytocin-vasopressinreceptor antagonist capable of inhibiting oxytocin-induceduterine contractions in both animals 20 andwomen with preterm labor. 21-24 The purpose of this studywas to evaluate the efficacy and safety of atosiban in thetreatment of patients in preterm labor.Material and methodsProtocol. Patients were eligible for participation if theymet the following criteria: preterm labor with intactmembranes, cervical dilatation of ≤3 cm, gestational ageof 20 weeks to 33 weeks 6 days, live fetus(es), and provisionof written informed consent. Gestational age was determinedby the best clinical estimate available. The diagnosisof preterm labor required the presence of ≥4uterine contractions over 30 minutes, each lasting at least40 seconds, and documented cervical change. The cervicalcriteria were met when either of the following waspresent: (1) a single cervical examination demonstratingdilatation of 1 cm to

Volume 182, Number 5 Romero et al 1175Am J Obstet GynecolFig 1. Disposition of patients enrolled. Asterisk, Medication errors. The majority of study drug given for patients whoreceived both was the assigned study drug.peutic failure (eg, 26 days vs 20 days). The estimates forthe mean number of days from the start of study drug todelivery or therapeutic failure were based on the resultsof an atosiban dose-ranging study 23 and the Canadianpreterm labor study. 6 Schoenfeld’s method 25 was usedfor the determination of sample size (α = .05).For efficacy, the primary intent-to-treat analyses wereperformed on the basis of all patients who received studydrug. The attrition rate in this study was minimal (6%);therefore from a clinical perspective the results from thepopulation of patients who received study drug are mostmeaningful. 26 However, intent-to-treat analyses includingall patients randomized were performed. The results ofthe intent-to-treat analyses of both populations led to thesame conclusion, but only the results for patients who receivedstudy drug are provided. The interval from start oftreatment to delivery or therapeutic failure was analyzedwith survival data analysis methods (log-rank test) andwas stratified by center. 27 Cox’s proportional-hazardsmodel was used for exploratory multivariate analysis forthis end point. Possible effects of alternate tocolyticsgiven for reasons other than progression of labor wereexcluded by censoring observations at the time suchagents were administered. Cochran-Mantel-Haenszeltests, stratified by center, were used to compare the percentagesof patients who remained undelivered and hadnot used an alternate tocolytic at 24 hours, 48 hours, and7 days. Logistic regression was used for exploratory multivariateanalyses of these 3 end points. All tests were2-sided with α = .05, with the exception of tests for interactions,which used α = .10. For the evaluation of safety,patients were classified according to the study drug theyactually received; patients who inadvertently receivedboth atosiban and placebo were placed in the atosibangroup. Maternal-fetal adverse events are summarized accordingto the WHOART dictionary. 28 Information onthe infants obtained at delivery is summarized with descriptivestatistics.For efficacy and safety end points, 2-sided 95% confidenceintervals for the difference in percentages ormeans between treatment groups are provided as descriptivestatistics. Summaries and results of analyses arepresented on the basis of the gestational age at admission.ResultsDisposition of patients and baseline characteristics ofthe population. Five hundred thirty-one patients were enrolledand randomized to treatment at 37 study centers,and 501 received the study drug. Fifteen patients randomizedto each treatment did not receive study drug forsimilar reasons. The most common reason was apost–randomization protocol exclusion that prohibitedstudy treatment of patients in whom cervical dilatationhad progressed beyond 3 cm before study drug could beprepared. Fig 1 displays a trial profile describing the dispositionof patients. The baseline demographic and clinicalcharacteristics of patients in both groups are displayedin Table I. The treatment groups werecomparable for most baseline variables including race,maternal age, and parity. However, the mean gestationalage on admission was statistically significantly greater forthe placebo group than for the atosiban group, and therewere nearly twice as many atosiban-treated patients enrolledat

1176 Romero et al May 2000Am J Obstet GynecolFig 2. Statistical curves for time to delivery interval or therapeutic failure in intent-to-treat subjects who received studydrug.Effect of atosiban on primary and secondary outcomes.The difference in time to delivery or therapeutic failurebetween atosiban and placebo was not statistically significant(median, 25.6 days; range, 0 to 115.5 days; vs 21.0days; range, 0 to 110.7 days, respectively; P = .6). This isreflected in survival curves for the time to delivery ortherapeutic failure (Fig 2). Cox’s proportional hazardsmodeling indicated a significant interaction (P < .10) betweentreatment and gestational age at entry, with resultsfavoring atosiban at higher gestational ages (≥28 weeks)and placebo at lower gestational ages (

Volume 182, Number 5 Romero et al 1177Am J Obstet GynecolTable I. Characteristics of study populationPlacebo Atosiban Statistical(n = 255) (n = 246) significanceGestational age at admissionMean and SD (wk) 31.0 (2.52) 30.3 (3.07) P = .008

1178 Romero et al May 2000Am J Obstet GynecolTable V. Clinical characteristics of pregnancies leading to fetal-infant deathPlacebo with or without rescue Atosiban with or without rescue Difference* and 95%(No. of deaths) (No. of deaths) confidence intervalGestational age at admission

Volume 182, Number 5 Romero et al 1179Am J Obstet GynecolTable VI. Summary of infant outcomes*Placebo with or without rescue Atosiban with or without rescue Difference and 95%End point (No.) (No.) confidence interval†Delivery‡

1180 Romero et al May 2000Am J Obstet GynecolFig 3. Disposition of patients according to response to placebo or active tocolysis. None of these 4 treatment subgroupsare comparable, and infant outcomes are expected to be different in each. Patients in premature labor allocatedto placebo may respond in either of two ways. Uterine quiescence may be achieved during placebo infusion.Many clinicians would interpret this as evidence that the patient was not in preterm labor (treatment subgroup 1).Alternatively, uterine contractility and cervical changes may continue, necessitating treatment with a tocolytic agent(treatment subgroup 2). These two treatment subgroups are clearly different because one group is in preterm laborand one is not. Patients allocated to atosiban (active) may also respond in either of two ways. Those who achieve uterinequiescence (treatment subgroup 3) are composed of a mixture of patients who are not in real preterm labor buthave a placebo response and those in real preterm labor who have responded to atosiban. In patients without a responseto atosiban (treatment subgroup 4) the initial trial of tocolysis failed.some of the adverse outcomes in infants exposed to atosibanmay be the result of an altered physiologic fetal responseto an insult. 33 Indeed, vasopressin secretion ispart of the fetal homeostatic response to stress. 34 Could apotential antivasopressin effect have contributed to thepoor outcomes among infants enrolled at

Volume 182, Number 5 Romero et al 1181Am J Obstet Gynecolsponse subgroups are examined. Indeed, the subgroupof patients with active treatment failure will always havethe worst perinatal outcomes (Fig 3). The complicationrate among infants in the atosiban failure subgroup (Fig3, subgroup 4) should be higher than that of the placebofailure subgroup (Fig 3, subgroup 2). With this study designthe treatment subgroups are not comparable and infantoutcomes will be different in each.Atosiban therapy was well tolerated in both the mothersand the fetuses. The side effect profiles of atosiban andplacebo were comparable (Table IV), with the exceptionof a higher incidence of injection-site reactions duringmaintenance among patients receiving atosiban. Most ofthe injection-site reactions that occurred during maintenancetherapy with the subcutaneous pump were minor.In some cases marked injection-site reactions led to discontinuationof therapy. The safety results associated withatosiban treatment compare favorably with publisheddata on the β-sympathomimetics, including ritodrine, theonly tocolytic agent approved in the United States. 38 Thesuperiority of atosiban maternal-fetal safety comparedwith that of ritodrine has also been demonstrated in 2randomized clinical trials. 23, 39 A significant difference betweenatosiban and β-sympathomimetics was observed inthe occurrence of pulmonary edema 40 (data also on filewith the R.W. Johnson Pharmaceutical ResearchInstitute) in the atosiban development program. Therewere no cases of pulmonary edema when atosiban was theonly tocolytic administered (0/1633), but there were 2cases of pulmonary edema when a β-sympathomimeticwas the only tocolytic administered (2/430).The current classes of tocolytic agents are far fromideal because all available agents have documented seriousside effects for either the mother or the infant, orboth, and some may have fetal toxicity. For example,there is a growing concern about the safety of magnesiumsulfate, probably the most widely used intravenoustocolytic agent used in the United States today. 41 Theseconcerns emanate from the results of the MAGNETtrial, 42, 43 as well as the Collaborative Eclampsia Trial ofmagnesium versus diazepam. 44 In this latter trial patientswith eclampsia treated with magnesium sulfate had aperinatal mortality that was twofold higher than that inpatients with eclampsia who were randomized to magnesiumand delivered before receiving the drug (25% vs12%). Other factors may be involved in the adverse outcomesreported in association with the administration ofmagnesium sulfate, but a toxic effect of magnesium cannotbe excluded.In this trial the treatment of patients in preterm laborwith atosiban resulted in prolongation of pregnancyfor up to 7 days for those at a gestational age ≥28 weekswith a low rate of maternal-fetal adverse effects. In addition,at a gestational age ≥28 weeks, infant morbidityand mortality of atosiban-initiated standard care weresimilar to those with placebo-initiated standard care.Given that all patients in this study were eligible for tocolysisand that, in practice, nearly all patients who areeligible for a tocolytic receive one, the benefit of usingatosiban is the placebo-like maternal-fetal side effectprofile. These observations support the short-term useof this oxytocin receptor antagonist in the treatment ofpatients in preterm labor with intact membranes.Efficacy and infant outcome data at

1182 Romero et al May 2000Am J Obstet GynecolAlan Golichowski, MD, PhD, Indiana UniversityMedical Center, Indianapolis, IndianaMarie Beall, MD, Harbor-University of California, LosAngeles Medical Center, Torrance, CaliforniaLawrence D. Devoe, MD, Medical College of Georgia,Augusta, GeorgiaDerek Wong, MD, Sutter Institute for MedicalResearch, Sacramento, CaliforniaMichael Paul, MD, Missouri Baptist Hospital, St Louis,MissouriMichael Pinette, MD, Maine Medical Center, Portland,MaineREFERENCES1. Report of the Consensus Development Conference on the effectof corticosteroids for fetal maturation on perinatal outcomes.Washington: National Institutes of Health; 1994 Nov. NIHPublication No.: 95-3784.2. Brown ER, Epstein MF. Neonatal consequences of pretermbirth. In: Fuchs A-R, Fuchs F, Stubblefield PG, editors. Pretermbirth: causes, prevention, and management. 2nd ed. New York:McGraw-Hill; 1993. p. 465-75.3. Escobedo MB. Follow-up of prematurely born infants. ClinObstet Gynecol 1988;31:662-70.4. Arias F, Tomich P. Etiology and outcome of low birth weight andpreterm infants. Obstet Gynecol 1982;60:277-81.5. King JF, Grant A, Keirse MJN. Betamimetics in preterm labour:an overview of the randomized controlled trials. Br J ObstetGynaecol 1988;95:211-22.6. The Canadian Preterm Labor Investigators Group. Treatment ofpreterm labor with the beta-adrenergic agonist ritodrine.N Engl J Med 1992;327:308-12.7. Merkatz IR, Peter JB, Borden TP. Ritodrine hydrochloride: abeta-mimetic agent for use in preterm labor. II. Evidence of efficacy.Obstet Gynecol 1980;56:712.8. Hudgens DR, Conradi SE. Sudden death associated with terbutalinesulfate administration. Am J Obstet Gynecol 1993;169:120-1.9. Katz M, Robertson PA, Creasy RK. Cardiovascular complicationsassociated with terbutaline treatment for preterm labor. Am JObstet Gynecol 1981;139:605-8.10. Foley MR, Landon MB, Gabbe SG, O’Dorisio TM, Waxman M,Leard R, et al. Effect of prolonged oral terbutaline therapy onglucose tolerance in pregnancy. Am J Obstet Gynecol 1993;168:100-5.11. Macones G, Sehdev H, Berlin M, Morgan M, Berlin J. Efficacy oforal beta-agonist maintenance therapy in preterm labor: a metaanalysis.Obstet Gynecol 1995;85:313-7.12. Besinger RE, Niebyl JR, Keyes WG, Johnson TR. Randomizedcomparative trial of indomethacin and ritodrine for the longtermtreatment of preterm labor. Am J Obstet Gynecol1991;164:981-8.13. Molnar M, Hertelendy F. Regulation of intracellular free calciumin human myometrial cells by prostaglandin F2 alpha:comparison with oxytocin. J Clin Endocrinol Metab 1990;71:1243-50.14. Norton ME, Merrill J, Cooper BAB, Kuller JA, Clyman RI.Neonatal complications after the administration of indomethacinfor preterm labor. N Engl J Med 1993;329:1602-7.15. Zingg H, Lefebvre D, Giaid A. Uterine oxytocin gene expression:a novel framework for oxytocin action. Regul Pept1993;45:43-6.16. Fuchs AR, Romero R, Deefe D, Parra M, Oyarzun E, Behnke E.Oxytocin secretion and human parturition: pulse frequency andduration increase during spontaneous labor in women. Am JObstet Gynecol 1991;165:1515-23.17. Fuchs AR, Fuchs F, Husslein P, Soloff MS. Oxytocin receptors inthe human uterus during pregnancy and parturition. Am JObstet Gynecol 1984;150:734-41.18. Soloff M, Fuchs AR, Fuchs F. Oxytocin receptors and the onset ofparturition. In: Albrecht E, Pepe G, editors. Perinatal endocrinology.Baltimore (MD): Perinatology Press; 1985. p. 289-311.19. Akerlund M, Stromberg P, Hauksson A, Andersen LF, LyndrupJ, Trojnar J, et al. Inhibition of uterine contractions of prematurelabour with an oxytocin analogue: results from a pilot study.Br J Obstet Gynaecol 1987;94:1040-4.20. Giussani DA, Jenkins SL, Mecenas CA, Winter JA, Barbera M,Honnebier OM, et al. The oxytocin antagonist atosiban preventsandrostenedione-induced myometrial contractions in thechronically instrumented, pregnant rhesus monkey.Endocrinology 1996;137:3302-7.21. Goodwin TM, Paul R, Silver H, Spellacy W, Parsons M, Chez R,et al. The effect of the oxytocin antagonist atosiban on pretermuterine activity in the human. Am J Obstet Gynecol 1994;170:474-8.22. Goodwin TM, Millar L, North L, Abrams L, Weglein R, HollandM. The pharmacokinetics of the oxytocin antagonist atosiban inpregnant women with preterm uterine contractions. Am JObstet Gynecol 1995;173:913-7.23. Goodwin TM, Valenzuela G, Silver H, Creasy G, the AtosibanStudy Group. Dose ranging study of the oxytocin antagonistatosiban in the treatment of preterm labor. Obstet Gynecol1996;88:331-61.24. Phaneuf S, Asboth G, MacKenzie IZ, Melin P, Lopez Bernal A.Effect of oxytocin antagonists on the activation of human myometriumin vitro: atosiban prevents oxytocin-induced desensitization.Am J Obstet Gynecol 1994;171:1627-34.25. Schoenfeld D. The asymptotic properties of nonparametric testsfor comparing survival distributions. Biometrika 1981;68:316-9.26. Gillings D, Koch G. The application of the principle of intention-to-treatto the analysis of clinicial trials. Drug Inform J1991;25:411-24.27. Lee ET. Statistical methods for survival data analysis. 2nd ed.New York: John Wiley; 1992. p. 1-4.28. WHOART version 1992, third quarter. WHO, The UppsalaMonitoring Center, WHO Collaborating Centre forInternational Drug Monitoring, Stora Torget 3, S-753 20Uppsala, Sweden.29. Goldenberg RL, Rouse DJ. Prevention of premature birth.N Engl J Med 1998;339:313-20.30. Keirse MNJC, Grant A, King JF. Preterm labour. In: Chalmers I,Enkin MW, Keirse MJNC, editors. Volume 1: effective care inpregnancy and childbirth. Oxford (United Kingdom): OxfordUniversity Press; 1989. p. 194-202.31. Romero R, Mazor M, Munoz H, Gomez R, Galasso M, ShererDM. The pretem labor syndrome. Ann N Y Acad Sci1994;734:414-29.32. Romero R, Sirtori M, Oyarzun E, Avila C, Mazor M, Callahan R,et al. Infection and labor. V. Prevalence, microbiology, and clinicalsignificance of intraamniotic infection in women withpreterm labor and intact membranes. Am J Obstet Gynecol1989;161:817-24.33. Gomez R, Romero R, Ghezzi F, Yoon BH, Mazor M, Berry SM.The fetal inflammatory response syndrome. Am J ObstetGynecol. 1998;179:194-202.34. Gore E. Perinatal fluid and electrolyte regulation: role of argininevasopressin. Semin Perinatol 1989;12:134-42.35. Advisory Committee for Reproductive Health Drugs. Bethesda(MD): Center for Drug Evaluation and Research, Food andDrug Administration; 1998 Apr 20. p. 198-9. New DrugApplication No.: 20-797.36. Advisory Committee for Reproductive Health Drugs. Pooledanalysis of the CAP-001 studies. Bethesda (MD): Center forDrug Evaluation and Research, Food and Drug Administration;1998 Apr 20. New Drug Application No.: 20-797.37. Advisory Committee for Reproductive Health Drugs. Bethesda(MD): Center for Drug Evaluation and Research, Food andDrug Administration; 1998 Apr 20. p. 24. New Drug ApplicationNo.: 20-797.38. Creasy RK, Katz M. Basic research and clinical experiences withβ-adrenergic tocolytics in the United States. In: Fuchs A-R,Fuchs F, Stubblefield PG, editors. Preterm birth: causes, preven-

Volume 182, Number 5 Romero et al 1183Am J Obstet Gynecoltion, and management. 2nd ed. New York: McGraw-Hill; 1993.p. 243-77.39. Moutquin J-M, Sherman D, Cohen H, Mohide PT, Hochner-Celnikier D, Fejgin M, et al. Double-blind, randomized, controlledtrial of atosiban and ritodrine in the treatment ofpreterm labor: a multicenter effectiveness and safety study. Am JObstet Gynecol 2000;182:1191-9.40. Advisory Committee for Reproductive Health Drugs. Bethesda(MD): Center for Drug Evaluation and Research, Food andDrug Administration; 1998 Apr 20. p. 60. New Drug ApplicationNo.: 20-797.41. Shankaran S, Papile LA, Wright LL, Ehrenkranz RA, Mele L,Lemons JA, et al. The effect of antenatal phenobarbital therapyon neonatal intracranial hemorrhage in preterm infants. N EnglJ Med 1997;337:466-71.42. Mittendorf R, Covert R, Bowman J, Khoshnood B, Lee KS,Siegler M. Is tocolytic magnesium sulphate associated with increasedtotal paediatric mortality? Lancet 1997;350:1517-8.43. Mittendorf R, Pryde P, Khoshnood B, Lee KS. If tocolytic magnesiumsulfate is associated with excess total pediatric mortality,what is its impact? Obstet Gynecol 1998;92:308-11.44. The Eclampsia Trial Collaborative Group. Which anticonvulsantfor women with eclampsia? Evidence from the CollaborativeEclampsia Trial. Lancet 1995;345:1455-63.

More magazines by this user
Similar magazines