Advertisement

Impact of a ‘rescue course’ of antenatal corticosteroids: a multicenter randomized placebo-controlled trial

      Objective

      Previous studies using repetitive courses of antenatal corticosteroids (ACS) have demonstrated marginal or no benefit and concern over potential risk. No prior prospective or randomized studies have evaluated the option of a single rescue course of ACS on neonatal outcome.

      Study Design

      A multicenter randomized double-blind placebo-controlled trial was performed from May 2003 through February 2008 in 18 private (15) and university (3) medical centers. Patients with singletons or twins < 33 weeks who had completed a single course of ACS before 30 weeks and at least 14 days before inclusion, and were judged to have a recurring threat of preterm delivery in the coming week, were included. Patients were randomized to receive a single rescue course of betamethasone, 2 12-mg doses 24 hours apart, or placebo. Exclusion criteria included: premature rupture of membranes, advanced dilation (> 5 cm), chorioamnionitis, and other steroid use.

      Results

      In all, 437 patients were randomized (223 rescue steroid group and 214 placebo group). A total of 55% of patients in each group delivered at < 34 weeks. There was a significant reduction in the primary outcome of composite neonatal morbidity < 34 weeks in the rescue steroid group vs placebo (43.9% vs 63.6%; odds ratio, 0.45; 95% confidence interval, 0.27-0.75; P = .002) and significantly decreased respiratory distress syndrome, ventilator support, and surfactant use. Perinatal mortality and other morbidities were similar in each group. Including all neonates in the analysis (regardless of gestational age at delivery) still demonstrated a significant reduction in composite morbidity in the rescue course group (32.1% vs 42.6%, odds ratio, 0.65; 95% confidence interval, 0.44-0.97; P = .0034) and improvement in respiratory morbidities.

      Conclusion

      Administration of a single rescue course of ACS before 33 weeks improves neonatal outcome without apparent increased short-term risk.

      Key words

      The administration of antenatal corticosteroids (ACS) to mothers who subsequently deliver prematurely is one of the few obstetric interventions that unequivocally improves outcome in premature babies.
      • Liggins G.C.
      • Howie R.N.
      A controlled trial of antepartum glucocorticoid treatment for prevention of the respiratory distress syndrome in premature infants.
      • Crowley P.
      Antenatal corticosteroid therapy: a metaanalysis of the randomized trials 1972-1994.
      Benefits to the newborn include reductions in mortality, respiratory distress syndrome (RDS), and intraventricular hemorrhage (IVH). One of the questions raised by the original randomized trials is the risk and benefit of repeating courses of ACS in patients who do not deliver in the week or 2 after the original course but who remain at risk of premature delivery. This question arose because of information suggesting that the beneficial effects of ACS were no longer apparent beyond a week after receiving the original ACS course.
      National Institutes of Health Consensus Conference Panel on the Effect of Corticosteroids for Fetal Maturation on Perinatal Outcomes.
      Subsequently a number of randomized controlled trials (RCTs) of multiple courses of ACS
      • Guinn D.A.
      • Atkinson M.W.
      • Sullivan L.
      • et al.
      Single vs weekly courses of antenatal corticosteroids for women at risk of preterm delivery: a randomized controlled trial.
      • Wapner R.J.
      • Sorokin Y.
      • Thom E.A.
      • et al.
      Single vs weekly courses of antenatal corticosteroids: evaluation of safety and efficacy; National Institute of Child Health and Human Development maternal fetal medicine units network.
      • Crowther C.A.
      • Haslam R.R.
      • Hiller J.E.
      • Doyle L.W.
      • Robinson J.S.
      Australian Collaborative Trial of Repeat Doses of Steroids (ACTORDS) Study Group
      Neonatal respiratory distress after repeat exposure to antenatal corticosteroids: a randomized controlled trail.
      were performed as well as neonatal follow-up studies to address this question.
      • French N.P.
      • Hagan R.
      • Evans S.F.
      • Godfrey M.
      • Newnham J.P.
      Repeated antenatal corticosteroids: size at birth and subsequent development.
      • Thorp J.A.
      • Jones P.G.
      • Knox E.
      • Clark R.H.
      Does antenatal corticosteroid therapy affect birth weight and head circumference?.
      • Crowther C.A.
      • Doyle L.W.
      • Haslam R.R.
      • et al.
      Outcomes at 2 years after repeat doses of antenatal corticosteroids.
      Although showing a modest reduction in RDS, concerns have been raised including the unnecessary treatment of many patients who did not deliver prematurely. In addition, decreased fetal growth and reduction in fetal head circumference in those babies receiving multiple doses of steroids has been shown in a number of studies evaluating repeated courses of ACS.
      • Wapner R.J.
      • Sorokin Y.
      • Thom E.A.
      • et al.
      Single vs weekly courses of antenatal corticosteroids: evaluation of safety and efficacy; National Institute of Child Health and Human Development maternal fetal medicine units network.
      • Crowther C.A.
      • Haslam R.R.
      • Hiller J.E.
      • Doyle L.W.
      • Robinson J.S.
      Australian Collaborative Trial of Repeat Doses of Steroids (ACTORDS) Study Group
      Neonatal respiratory distress after repeat exposure to antenatal corticosteroids: a randomized controlled trail.
      • French N.P.
      • Hagan R.
      • Evans S.F.
      • Godfrey M.
      • Newnham J.P.
      Repeated antenatal corticosteroids: size at birth and subsequent development.
      • Thorp J.A.
      • Jones P.G.
      • Knox E.
      • Clark R.H.
      Does antenatal corticosteroid therapy affect birth weight and head circumference?.
      The National Institutes of Health (NIH), in its second consensus conference regarding ACS, and the opinion of the American College of Obstetricians and Gynecologists, concluded that “repetitive doses of ACS could not be recommended outside of clinical trials including so-called rescue doses.”
      National Institutes of Health Consensus Development Panel
      ACOG committee opinion no. 402: antenatal corticosteroid therapy for fetal maturation.
      The concept of a rescue course of ACS basically involves administering a second course to patients whose pregnancies continue more than a week or 2 from their original course and only in whom, in the judgment of the clinician, delivery has again become likely. This is the 1 approach that has not yet been evaluated in any of the prospective randomized trials of ACS.
      For Editors' Commentary, see Table of Contents
      See related editorial, page 217
      We, therefore, chose to perform a multicenter trial to evaluate the potential benefit of a single rescue course of ACS to answer this last major unanswered question regarding the optimal method of ACS administration.

      Materials and Methods

      We performed a multicenter randomized double-blind placebo-controlled trial to evaluate the impact of a rescue course of antenatal steroids on the incidence of neonatal morbidity and mortality. Study approval was obtained from every participating hospital's institutional review board and all women signed written informed consent. Eligible women with singleton or twin pregnancies with intact membranes at 25-32 6/7 weeks, judged to have a recurrent or continued threat of preterm delivery within the next 7 days, and who initiated their first course of betamethasone at least 14 days before enrollment and before 30 weeks' gestation, were consented and enrolled.
      Consenting patients were randomized to either a repeated course of betamethasone 12 mg intramuscularly, 2 doses given 24 hours apart, or placebo. The research pharmacist (unblinded) at each site prepared the medication based on a blocked randomization sequence that was prepared centrally. The syringes were completely covered by a label to conceal the contents. As betamethasone had become unavailable in many US hospitals, in such a situation the protocol allowed for the alternative of dexamethasone to be used in a dose of 6 mg intramuscularly every 12 hours × 4 doses total, with the placebo group receiving the same volume and frequency of normal saline. Patients excluded were those with known major fetal anomalies, with high-order multiple gestation (≥ triplets), with cervical dilation ≥ 5 cm, with ruptured membranes, with clinical chorioamnionitis, with documented lung maturity, receiving corticosteroids for other maternal indications, and with human immunodeficiency virus or active tuberculosis. The remainder of clinical care remained at the discretion of the clinician.
      The primary outcome measure, as specified by the protocol, was composite neonatal morbidity in babies delivering < 34 weeks. Composite morbidity was defined as ≥ 1 of the following: RDS (oxygen requirement, clinical diagnosis, and consistent chest radiograph), bronchopulmonary dysplasia (requirement for oxygen support at 30 days of life), severe IVH (grades III or IV), periventricular leukomalacia, blood culture-proven sepsis, necrotizing enterocolitis, or perinatal death (stillbirth or death before neonatal hospital discharge). Secondary outcome measures included: preterm delivery before 34 weeks, RDS alone, gestational age at delivery, birth weight, intrauterine growth restriction (IUGR), head circumference, need for surfactant therapy, pneumothorax, and maternal infectious morbidity.

      Statistical Tests and Analyses

      We hypothesized that administration of a second rescue course of ACS, compared with 1 course, would show a 40% reduction in the incidence of composite neonatal outcome in patients delivering before 34 weeks' gestation. The sample size was estimated based on a composite morbidity of 28%.
      • Guinn D.A.
      • Atkinson M.W.
      • Sullivan L.
      • et al.
      Single vs weekly courses of antenatal corticosteroids for women at risk of preterm delivery: a randomized controlled trial.
      Each arm required at least 217 subjects to have 80% power to detect a 40% reduction to 16.8% (2-tailed, alpha = .05) using a comparison for proportions between groups (Fisher exact test).
      Although the prespecified primary outcome was composite neonatal morbidity/mortality in patients delivering before 34 weeks' gestation, analyses were also conducted in all randomized women with a known outcome and in the treatment group to which they were randomized (modified intent to treat). Analysis of the primary outcome used a repeated measures approach (generalized estimating equations with an exchangeable correlation structure
      • Liang K.Y.
      • Zeger S.L.
      Longitudinal data analysis using generalized linear models.
      where each baby was considered the repeated measure). Odds ratios, 95% confidence intervals, and P values were determined from the repeated measures model.
      Differences between treatment groups in baseline maternal characteristics and maternal delivery characteristics were determined using a 2-tailed χ2 test and Fisher exact test for dichotomous variables and a Wilcoxon rank sum test for continuous or ordinal variables. Differences between the treatment groups in neonatal characteristics and secondary outcome measures were determined from a repeated measures model where each baby was considered the repeated measure (generalized estimating equations for dichotomous variables and a mixed linear model for continuous variables). All analyses were performed using software (SAS 9.1.3; SAS Institute Inc, Cary, NC).
      An internal safety monitoring committee, which included 3 individuals not participating in the study, reviewed safety and efficacy data twice while the study was ongoing and reviewed all serious adverse events. Two interim analyses of the primary efficacy outcome were conducted when 24.3% (140/577) and 52.3% (302/577) of babies were delivered. Thus, based on the O'Brien-Fleming spending function, the adjusted alpha level for the primary outcome is 0.049. For all other analyses, no adjustments are made and an alpha level of 0.05 is used. All statistical tests appear as 2-sided P values.

      Results

      This multicenter study was performed from May 2003 through February 2008 in 18 private (15) and university (3) medical centers and conducted by investigators from 10 Obstetrix/Pediatrix Medical Group (Sunrise, FL) maternal-fetal medicine practices and 4 other practices staffed by maternal-fetal medicine attending physicians. The population was largely private/nongovernment-funded patients. After screening a total of 2122 patients for eligibility (Figure), 437 were consented and randomized, 223 to the ACS group and 214 to the placebo group. There were a total of 289 babies in the ACS group (67 women were pregnant with twins, although 1 woman had a fetal death before enrollment in the study) and 288 babies in the placebo group (74 women had twins). Neonatal data were missing for 12 and 6 babies in the ACS and placebo groups, respectively, resulting in 276 and 282 babies available for analysis of the primary efficacy outcome.
      Figure thumbnail gr1
      FIGUREConsort Flow Diagram
      Overview of clinical trail enrollment.
      ACS, antenatal corticosteroids; OR, odds ratio.
      Garite. Impact of a ‘rescue course’ of antenatal corticosteroids: a multicenter randomized placebo-controlled trial. Am J Obstet Gynecol 2009.
      In all, 196 (88%) of the ACS group and 184 (86%) of the placebo group received a full course of antenatal steroid or placebo. In all, 31 patients received dexamethasone and 30 received the dexamethasone placebo of which 23 and 22 received a full course, respectively.
      As the protocol specified, primary outcome of the study was neonatal composite morbidity at < 34 weeks; there were 124 (56%) ACS and 119 (55%) placebo patients who actually delivered at < 34 weeks. The mean intervals between randomization and delivery were 24.5 and 25.1 days, respectively. Premature delivery (< 36 weeks) was also similar between the 2 groups: ACS (168/218; 77.1%) and placebo (173/212; 81.6%). At enrollment and at delivery the groups were similar in all evaluated variables (TABLE 1, TABLE 2).
      TABLE 1Baseline characteristics
      VariablesAll patientsGA at delivery < 34 wk
      ACSPlaceboACSPlacebo
      No. in each group223214124119
      Maternal age (y), mean ± std29 ± 629 ± 629 ± 628 ± 5
      Maternal height (in), mean ± std64 ± 364 ± 364 ± 364 ± 3
      Maternal weight (lb), mean ± std157 ± 43160 ± 47158 ± 46.160 ± 42
      GA (wk) at randomization, mean ± SD29.5 ± 2.229.4 ± 1.929.1 ± 2.329.1 ± 1.9
      Parity, n (%)
       096 (43)84 (39.3)56 (45.1)49 (41.1)
       165 (29.1)76 (35.5)34 (27.4)41 (34.5)
       ≥ 262 (27.8)54 (25.2)34 (27.4)29 (24.4)
      Married, n (%)141 (63.2)142 (66.4)78 (62.9)77 (64.7)
      Ethnicity, n (%)
       African American12 (5.4)21 (9.8)9 (7.3)12 (10.1)
       Asian or Pacific Islander14 (6.3)12 (5.6)10 (8.1)10 (8.4)
       Caucasian133 (59.6)114 (53.3)71 (57.3)53 (44.5)
       Hispanic51 (22.9)57 (26.6)29 (23.4)40 (33.6)
       Native American6 (2.7)4 (1.9)2 (1.6)2 (1.7)
       Other7 (3.1)6 (2.8)3 (2.4)2 (1.7)
      Maternal GBS, n (%)28 (12.6)28 (13.1)16 (12.9)16 (13.5)
      Smoker, n (%)36 (16.1)38 (17.8)24 (19.4)19 (16.0)
      ETOH abuse, n (%)3 (1.3)4 (1.9)3 (2.4)1 (0.8)
      Report of drug abuse, n (%)10 (4.5)15 (7)7 (5.7)7 (5.9)
      Twin pregnancy, n (%)67 (30)74 (34.6)44 (35.5)47 (39.5)
      ACS, antenatal corticosteroids; ETOH, alcohol; GA, gestational age; GBS, group B Streptococcus; SD, standard deviation.
      Garite. Impact of a ‘rescue course’ of antenatal corticosteroids: a multicenter randomized placebo-controlled trial. Am J Obstet Gynecol 2009.
      TABLE 2Comparison of groups at delivery
      No. in each groupAll patients34 wk at delivery
      ACSPlaceboPACSPlaceboP
      Enrolled mothers223214124119
      Fetuses/newborns (each baby counted in twin pairs)289288167166
      Report of tocolysis, n (%)181 (81.2)173 (80.8).93199 (79.8)92 (77.3).631
      Antibiotics reported, n (%)149 (66.8)139 (65.0).68178 (62.9)82 (68.9).324
      Gestational age (wk) at delivery, mean (SD)33.1 ± 3.133.0 ± 3.1.87730.9 ± 2.030.8 (1.8).705
      Birth weight (g), mean (SD)1905 ± 7381920 ± 667.8051517 ± 4781506 ± 413.861
      Head circumference (cm)30.2 ± 3.030.0 ± 2.9.52328.5 ± 2.528.6 ± 2.7.760
      Delivery reason, n (%)
       Maternal hemorrhage/abruption9 (3.1)11 (3.8).6436 (3.6)6 (3.6).989
       Amniocentesis for +FLM or term20 (6.9)34 (11.8).10604 (2.4)
       Clinical chorioamnionitis2 (0.7)2 (0.7).9972 (1.2)2 (1.2).995
       Fetal distress4 (1.4)16 (5.6).0223 (1.8)13 (7.8).041
       Incompetent cervix1 (0.4)4 (1.4).1671 (0.6)4 (1.2).170
       Maternal complication15 (5.2)14 (4.9).90412 (7.2)6 (3.6).205
       Nonreassuring fetal testing20 (6.9)17 (5.9).39916 (9.6)16 (9.6).635
       Oligohydramnios3 (1.0)4 (1.4).7682 (1.2)0NC
       PIH14 (4.8)18 (6.3).74210 (6.0)16 (9.6).404
       Preterm labor unstoppable or not stopped89 (30.1)93 (32.3).55453 (31.7)59 (35.5).297
       PROM infection1 (0.4)0NC1 (0.6)0NC
       PROM/labor35 (12.1)25 (8.7).329025 (15.0)17 (10.2).358
       Severe IUGR7 (2.4)6 (2.1).7425 (3.0)1 (0.6).100
       Worsening fetal condition14 (4.8)11 (3.8).66511 (6.6)10 (6.0).831
       Other43 (14.9)27 (9.4).07017 (10.2)10 (6.0).159
       Unknown12 (4.2)5 (1.7).1253 (1.8)1 (0.6).574
      Route of delivery, n (%)
       Cesarean section186 (64.4)202 (70.1).244119 (71.3)132 (79.5).211
       Spontaneous vaginal delivery80 (27.7)76 (26.4).69240 (24.0)31 (18.7).350
       Operative vaginal delivery9 (3.1)4 (1.4).2584 (2.4)1 (0.6).378
       Unknown (lost to follow-up)14 (4.8)6 (2.1).1454 (2.4)2 (1.2).622
      Cord pH arterial median7.27 ± .077.28 ± .08.7537.29 ± .077.27 ± .09.128
      Cord pH venous median7.33 ± .077.32 ± .08.2777.34 ± .077.32 ± .08.062
      Apgar 1 min < 5, n (%)32 (11.1)43 (14.9).16627 (16.2)33 (19.9).274
      Apgar 5 min < 7, n (%)18 (6.2)23 (8).33817 (10.2)18 (10.8).636
      ACS, antenatal corticosteroids; FLM, fetal lung maturity; IUGR, intrauterine growth restriction; NC, not calculable; PIH, pregnancy induced hypertension; PROM, premature rupture of membranes; SD, standard deviation.
      Garite. Impact of a ‘rescue course’ of antenatal corticosteroids: a multicenter randomized placebo-controlled trial. Am J Obstet Gynecol 2009.
      The maternal outcomes did not differ between the 2 groups. In particular, rates of chorioamnionitis (6 vs 9) and sepsis (2 vs 1) were similar for the ACS vs placebo groups, respectively. There was 1 maternal death in a placebo group patient who withdrew from the study before receiving any injection.
      For the primary endpoint, there was a significant reduction in composite morbidity at < 34 weeks in the ACS group vs placebo group (Table 3). In addition, at this gestational age in the ACS group there were fewer cases of RDS, less need for surfactant, and less frequent need for ventilator support. There were no differences in mortality or any of the other morbidities at this gestational age including IVH, sepsis, and pneumonia. Including all babies at all gestational ages, there was also a statistically significant reduction in composite morbidity, RDS, and use of surfactant, but not in need for ventilator therapy; and there were no differences in the remainder of the morbidities or mortality. The rates of perinatal death were also similar (Table 4). There was 1 stillbirth and 4 neonatal deaths in the ACS group and 1 stillbirth and 6 neonatal deaths in the placebo group. Of the neonatal deaths, 3 in the ACS group and 4 in the placebo group were related to complications of prematurity (IVH, necrotizing enterocolitis, sepsis).
      TABLE 3Neonatal morbidity and mortality-delivery < 34 weeks
      Neonatal morbidityACS n/N (%)Placebo n/N (%)OR
      OR, 95% CI, and P value are determined using a repeated measures model where each twin is considered a repeated measure.
      (95% CI)
      P value
      OR, 95% CI, and P value are determined using a repeated measures model where each twin is considered a repeated measure.
      Composite morbidity71/163 (43.9)105/165 (63.6)0.45 (0.27-0.75).002
      RDS67/162 (41.4)101/164 (61.6)0.45 (0.27-0.75).002
      BPD27/160 (16.9)20/163 (12.3)1.53 (0.77-3.07).228
      Surfactant61/162 (37.7)91/164 (55.5)0.49 (0.30-0.80).004
      Ventilator59/157 (37.6)83/157 (52.9)0.56 (0.33-0.92).023
      ROP21/153 (13.7)19/151 (12.6)0.12 (0.54-2.34).760
      ROP grade 3 or 43/153 (2.0)2/151 (1.3)2.47 (0.25-24.58).442
      IVH17/160 (10.6)23/163 (14.1)0.72 (0.34-1.50).378
      IVH grade 3 or 45/160 (3.1)4/162 (2.5)1.34 (0.32-5.57).684
      NEC8/162 (4.9)11/163 (6.8)0.70 (0.26-1.90).485
      Sepsis7/162 (4.3)12/164 (7.3)0.54 (0.19-1.58).262
      Perinatal death5/162 (3.1)7/166 (4.2)0.61 (0.17-2.21).451
      Pneumothorax3/162 (1.9)4/164 (2.3)0.77 (0.17-3.45).734
      PVL3/159 (1.9)4/161 (2.5)0.76 (0.17-3.41).720
      Pneumonia3/161 (1.9)5/164 (3.1)0.64 (0.14-3.01).577
      IUGR (< 10th percentile)25/162 (15.4)20/161 (12.4)1.33 (0.70-2.55).383
      ACS, antenatal corticosteroids; BPD, bronchopulmonary dysplasia; CI, confidence interval; IUGR, intrauterine growth restriction; IVH, intraventricular hemorrhage; NEC, necrotizing enterocolitis; OR, odds ratio; PVL, periventricular leukomalacia; RDS, respiratory distress syndrome; ROC, retinopathy of prematurity.
      Garite. Impact of a ‘rescue course’ of antenatal corticosteroids: a multicenter randomized placebo-controlled trial. Am J Obstet Gynecol 2009.
      a OR, 95% CI, and P value are determined using a repeated measures model where each twin is considered a repeated measure.
      TABLE 4Neonatal morbidity–all newborns
      Neonatal morbidityACS n/N (%)Placebo n/N (%)OR
      OR, 95% CI, and P value are determined using a repeated measures model where each twin is considered a repeated measure.
      (95% CI)
      P value
      OR, 95% CI, and P value are determined using a repeated measures model where each twin is considered a repeated measure.
      Composite morbidity88/276 (32.1)120/282 (42.6)0.65 (0.44-0.97).034
      RDS83/275 (30.2)116/281 (41.3)0.64 (0.43-0.95).026
      BPD27/273 (9.9)20/278 (7.2)1.50 (0.76-2.94).239
      Surfactant70/273 (25.6)99/280 (35.4)0.65 (0.43-0.98).038
      Ventilator70/267 (26.2)95/273 (34.8)0.70 (0.46-1.06).088
      ROP21/260 (8.1)19/267 (7.1)1.15 (0.56-2.34).704
      ROP grade 3 or 43/260 (1.2)2/267 (0.8)2.51 (0.25-24.83).432
      IVH19/272 (7.0)25/274 (9.1)0.76 (0.38-1.49).422
      IVH grade 3 or 46/272 (2.2)4/274 (1.5)1.65 (0.43-6.31).462
      NEC8/275 (2.9)11/280 (3.9)0.72 (0.27-1.91).503
      Sepsis7/275 (2.6)12/288 (4.3)0.55 (0.19-1.60).275
      Perinatal death5/276 (1.8)7/282 (2.5)0.60 (0.17-2.18).440
      Pneumothorax4/275 (1.5)4/281 (1.4)1.03 (0.26-4.13).967
      PVL3/269 (1.1)4/272 (1.5)0.76 (0.17-3.40).718
      Pneumonia3/274 (1.1)5/281 (1.8)0.66 (0.14-3.04).589
      IUGR (< 10th percentile)29/275 (10.6)28/278 (10.1)1.07 (0.61-1.86).822
      ACS, antenatal corticosteroids; CI, confidence interval; IUGR, intrauterine growth restriction; IVH, intraventricular hemorrhage; NEC, necrotizing enterocolitis; OR, odds ratio; RDS, respiratory distress syndrome.
      Garite. Impact of a ‘rescue course’ of antenatal corticosteroids: a multicenter randomized placebo-controlled trial. Am J Obstet Gynecol 2009.
      a OR, 95% CI, and P value are determined using a repeated measures model where each twin is considered a repeated measure.
      We also analyzed the questions regarding the timing and duration of effectiveness of this second rescue course of ACS. The largest and most significant difference in composite morbidity was seen between 2 and 7 days elapsed from the first dose of the rescue course (Table 5). Although not a predesignated analysis, it is also clinically important to examine in which gestational ages the greatest efficacy was seen. The reduction in composite morbidity with ACS was limited to babies < 33 weeks: 62/129 (48%) ACS vs 95/131 (72.5%) placebo (odds ratio, 0.40; 95% confidence interval, 0.23-0.71; P < .002). There was no difference in outcome at ≥ 33 weeks: 26/147 (18%) ACS vs 25/155 (17%) placebo (P = .811). These 2 groups were otherwise similar in gestational age at randomization and latency to delivery and other entry variables.
      TABLE 5Neonatal composite morbidity by days from first dose to delivery
      ACS n/N (%)Placebo n/N (%)P value
      Odds ratio, 95% confidence interval, and P value are determined using a repeated measures model where each twin is considered a repeated measure.
      0 or 1 d13/25 (52.0)15/20 (75.0).067
      2-7 d24/61 (39.3)37/53 (69.8).035
      8-14 d15/37 (40.5)24/47 (51.5).328
      15-21 d9/22 (40.9)21/32 (65.6).173
      22-28 d13/31 (41.9)12/39(30.8).550
      > 28 d14/100 (14.0)11/91 (12.1).588
      ACS, antenatal corticosteroids.
      Garite. Impact of a ‘rescue course’ of antenatal corticosteroids: a multicenter randomized placebo-controlled trial. Am J Obstet Gynecol 2009.
      a Odds ratio, 95% confidence interval, and P value are determined using a repeated measures model where each twin is considered a repeated measure.
      Neonatal anthropomorphic measurements are of particular interest, and there was no difference in birth weight, rates of IUGR, or head circumference (Table 2). As lagging body or head growth due to repetitive corticosteroids may not be reflected in babies who deliver soon after treatment, we also analyzed these parameters in babies delivering after 36 weeks and > 14 days after treatment (n = 49 ACS and 45 placebo). This subanalyses demonstrated similar mean head circumferences (33.0 cm study vs 32.5 cm placebo), mean birth weights (2821g vs 2803 g), and rates of IUGR (0 vs 1; 2.2%). In neither analysis was there a statistical difference or a trend toward reduced body or head growth in rescue steroid–treated pregnancies; however, the sample size of this analysis may not be sufficient to detect even modest differences.

      Comment

      There are 2 important clinical questions addressed by this study. The first is whether the use of rescue steroids reduces morbidity and/or mortality in patients who have been previously treated with ACS but who again threaten to deliver before 34 weeks. The results of this study demonstrate a benefit in composite morbidity and decrease in the presence and severity of RDS, but not mortality or other morbidities at < 34 weeks. This benefit was also seen in the overall group, but of course this is affected by the magnitude of benefit in the earlier gestations and the fact that the majority delivered before 34 weeks.
      The second important question addressed is whether this discretionary method of ACS administration based on the clinician's judgment of impending risk as compared with simply giving repeated doses to all patients at risk (eg, all twins, all previous preterm deliveries) more correctly identifies babies who will actually deliver before 34 weeks (and, therefore, would potentially benefit by treatment). This is the only prospective and randomized study to date evaluating this rescue approach to ACS treatment. One retrospective study by Vermillion et al
      • Vermillion S.T.
      • Bland M.L.
      • Soper D.E.
      Effectiveness of a rescue dose of antenatal betamethasone after an initial single dose.
      did demonstrate a reduction in RDS using a rescue approach, but that study also differed in that a single dose of betamethasone was used rather than a full 2-dose course. A previous prospective randomized study by Peltoniemi et al
      • Peltoniemi O.M.
      • Kari M.A.
      • Tammela O.
      • et al.
      Randomized trial of a single repeat dose of prenatal betamethasone treatment in imminent preterm birth.
      also used a single booster dose of betamethasone in patients, but showed no benefit, as 79% of their patients delivered in < 24 hours. In contrast, 89% of our patients received a full rescue course of ACS, and 55% of patients did deliver before 34 weeks. Overall, 78% of our patients delivered before 36 weeks. Of the previous multicenter studies that used routine repetitive doses of ACS, 2 of the 3 included patients with preterm premature rupture of membranes (PPROM), who are obviously at much higher risk of imminent delivery than the patients included in this study.
      • Guinn D.A.
      • Atkinson M.W.
      • Sullivan L.
      • et al.
      Single vs weekly courses of antenatal corticosteroids for women at risk of preterm delivery: a randomized controlled trial.
      • Crowther C.A.
      • Haslam R.R.
      • Hiller J.E.
      • Doyle L.W.
      • Robinson J.S.
      Australian Collaborative Trial of Repeat Doses of Steroids (ACTORDS) Study Group
      Neonatal respiratory distress after repeat exposure to antenatal corticosteroids: a randomized controlled trail.
      The only other previous study that also excluded patients with PPROM had only 36% of patients who delivered before 34 weeks and 56% who delivered before 36 weeks.
      • Wapner R.J.
      • Sorokin Y.
      • Thom E.A.
      • et al.
      Single vs weekly courses of antenatal corticosteroids: evaluation of safety and efficacy; National Institute of Child Health and Human Development maternal fetal medicine units network.
      Therefore, it appears with this discretionary method, as opposed to routine repetitive dosing, we were better able to correctly identify patients who would truly benefit by this treatment and avoid unnecessarily treating those who would not (ie, deliver at or near term).
      In clinical practice the potential downside to this rescue steroid approach, however, is the question of how many babies who might have benefited by a routine repetitive dosing approach may not have been treated due to delivery shortly after admission. This issue could not be answered by this study as there was no opportunity to enroll or randomize patients at or immediately after their first course of ACS because many patients were maternal transfers who had received the course previously at another institution. However, only 38 (9%) of our patients delivered before receiving a complete rescue course of ACS and only 10 (2%) delivered before receiving the first dose of the rescue course. The NIH consensus conference on ACS
      National Institutes of Health Consensus Conference Panel on the Effect of Corticosteroids for Fetal Maturation on Perinatal Outcomes.
      concluded that even delivering 24 hours after the first dose of steroids imparts benefit, although this is only proven with the first course. Thus, it is possible that many of these 38 babies in our study who delivered before a full rescue course of ACS may have derived some benefit. The answer to whether a routine second course of steroids as opposed to a rescue course is the more optimal method could and perhaps should only be answered in a randomized trial comparing the 2 methods.
      From a purely scientific standpoint this is not a true intent-to-treat study as the primary endpoint was reduction in morbidity in patients who delivered at < 34 weeks as opposed to the entire population enrolled in the study. Technically, it could be said that this is a subgroup analysis. This argument is mitigated by the fact that the benefit in both composite and respiratory morbidity is seen in the entire study group including all patients who delivered before and after 34 weeks. We are compelled to use this primary endpoint of < 34 weeks as this was the actual hypothesis of the study. This endpoint was chosen for 2 reasons. First, virtually all studies have shown respiratory benefit of ACS at < 34 weeks. And the second question we wanted to answer was whether such a selective approach actually better identifies those patients who will deliver < 34 weeks as opposed to the routine repetitive approach, the downside of which is that more patients are unnecessarily treated, ultimately delivering beyond 34 weeks.
      We chose not to include patients with PPROM in this study for 2 reasons. Two previous studies have shown an increase in serious maternal and neonatal infectious morbidity in patients receiving a second course of ACS in the setting of ruptured membranes.
      • Vermillion S.T.
      • Soper D.E.
      • Chasedunn-Roark J.
      Neonatal sepsis after betamethasone to patients with preterm premature rupture of membranes.
      • Lee M.
      • Davies J.
      • Guinn D.
      • et al.
      Single vs weekly course of antenatal corticosteroids in preterm premature rupture of membranes.
      In addition, the efficacy of ACS in studies of patients with ruptured membranes is not as consistently demonstrated as compared with those studies of patients with intact membranes.
      National Institutes of Health Consensus Conference Panel on the Effect of Corticosteroids for Fetal Maturation on Perinatal Outcomes.
      One previous RCT of multiple courses of ACS and follow-up studies on babies in previous ACS trials suggested that multiple courses of ACS decreased overall fetal growth and the growth of the fetal brain as reflected in head circumference.
      • Wapner R.J.
      • Sorokin Y.
      • Thom E.A.
      • et al.
      Single vs weekly courses of antenatal corticosteroids: evaluation of safety and efficacy; National Institute of Child Health and Human Development maternal fetal medicine units network.
      • French N.P.
      • Hagan R.
      • Evans S.F.
      • Godfrey M.
      • Newnham J.P.
      Repeated antenatal corticosteroids: size at birth and subsequent development.
      • Thorp J.A.
      • Jones P.G.
      • Knox E.
      • Clark R.H.
      Does antenatal corticosteroid therapy affect birth weight and head circumference?.
      The 2-year follow-up study of Crowther et al,
      • Crowther C.A.
      • Doyle L.W.
      • Haslam R.R.
      • et al.
      Outcomes at 2 years after repeat doses of antenatal corticosteroids.
      however, did not show any differences with multiple doses of corticosteroids except that more babies in the steroid group required medical attention at 2 years than in the placebo group. However, in the studies that did suggest a deleterious effect, minimal or no apparent effects were seen until ≥3 courses of ACS were administered. Although acknowledging that our study was not powered to address these safety issues, we did not observe a difference or a trend toward reduction in birth weight or head circumference in the ACS-treated group including the subanalysis of just those babies where sufficient time elapsed to reflect this lack of growth was allowed. A recent follow-up study of babies from the National Institute of Child Health and Human Development randomized trial who received multiple course of ACS showed similar neurologic developmental testing scores, and similar growth. There was, however, a nonstatistically significant increase in the number of babies with cerebral palsy (6 ACS vs 1 in control) but 5 of 6 babies with cerebral palsy in the ACS group had been exposed to ≥ 4 courses of ACS.
      • Wapner R.J.
      • Sorokin Y.
      • Mele L.
      • et al.
      Long term outcomes after repeat doses of antenatal corticosteroids.
      Nonetheless, given the suggestions of potential harm with more frequent courses of ACS, it would seem prudent to limit the absolute number of courses given and to limit treatment to those babies most likely to deliver at a gestational age where they would benefit from ACS treatment. Thus, the rescue approach to retreatment used in this study would be most likely to achieve these goals.
      We, therefore, conclude that choosing to administer a rescue course of ACS in pregnant women treated initially > 2 weeks prior, and who are judged by the clinician to be likely to deliver within the next week and before 34 weeks of gestation, is a beneficial approach that significantly decreases respiratory complications of prematurity and is without apparent immediate or short-term adverse effects to the mother or baby.

      Acknowledgments

      Tabled 1
      Practice/hospital sitePrincipal investigatorResearch nurse(s)
      • Good Samaritan Hospital
      • San Jose CA
      Andrew Combs, MDKimberly Ireland, RN
      • Phoenix Perinatal Associates, Obstetrix
      • Banner Good Samaritan Medical Center
      • Banner Desert Samaritan Medical Center
      • Phoenix, AZ
      Garrett Lam, MD
      • Melissa Ingersoll, RN
      • Ana Braescu, MSN
      • Eastside Maternal–Fetal Medicine, Obstetrix
      • Evergreen Hospital
      • Kirkland, WA
      Martin Walker, MDSally Ann de-Vitry Smith, RN
      • Fetal Diagnostic Center, Obstetrix
      • Saddleback Memorial Medical Center
      • Laguna Hills, CA
      James Kurtzman, MDSandy Grose, RN
      • Obstetrix Medical Group of Washington
      • Swedish Medical Center
      • Seattle, WA
      David Luthy, MDTina Lopez, RN
      • Kansas City Obstetrix
      • St. Luke's Hospital
      • Kansas City, MO
      George Lu, MDMary Jean Brown, RN
      • Obstetrix Medical Group of Arizona
      • Tucson Medical Center
      • Tucson AZ
      Hugh Miller, MDDiane Mercer, RN
      • Southern California Obstetrix
      • Long Beach Memorial Medical Center
      • Long Beach, CA
      Michael Nageotte, MD
      • Christine Preslicka, RN
      • Bethany Vinson, RN
      • Deysi Caballero, LVN
      • University of California Irvine
      • Orange, CA
      Manuel Porto, MD
      • Pamela Rumney, RN
      • Lizette Spiers
      • Obstetrix Medical Group of Colorado
      • Presbyterian St. Lukes Medical Center
      • Swedish Medical Center
      • Rose Medical Center
      • Denver, CO
      Robert Stettler, MD
      • Leslie Harden, MSN
      • Mari Gambotto, PNNP
      • Maryanne Bruno, PNP
      • Marilyn Hall, CNM
      • Gwen Dubois, RN
      • Erlanger Medical Center
      • Chattanooga, TN
      • University of Tennessee Medical Center
      • Knoxville, TN
      C. David Adair, MD
      • Lorrie Mason, MSN
      • Stephanie Frazer, MSN
      • Tufts New England Medical Center
      • Boston, MA
      • Michael House, MD
      • Adam Wolfberg, MD
      • St. John's Regional Health Center
      • Springfield, MO
      Robert Fraser, MDCarla Weber, RN
      • Sunrise Hospital
      • Las Vegas NV
      Brian Iriye, MDJudy Hancock, RNC

      References

        • Liggins G.C.
        • Howie R.N.
        A controlled trial of antepartum glucocorticoid treatment for prevention of the respiratory distress syndrome in premature infants.
        Pediatrics. 1972; 50: 515-525
        • Crowley P.
        Antenatal corticosteroid therapy: a metaanalysis of the randomized trials 1972-1994.
        Am J Obstet Gynecol. 1995; 173: 322-335
      1. National Institutes of Health Consensus Conference Panel on the Effect of Corticosteroids for Fetal Maturation on Perinatal Outcomes.
        JAMA. 1995; 273: 413-418
        • Guinn D.A.
        • Atkinson M.W.
        • Sullivan L.
        • et al.
        Single vs weekly courses of antenatal corticosteroids for women at risk of preterm delivery: a randomized controlled trial.
        JAMA. 2001; 286: 1581-1587
        • Wapner R.J.
        • Sorokin Y.
        • Thom E.A.
        • et al.
        Single vs weekly courses of antenatal corticosteroids: evaluation of safety and efficacy; National Institute of Child Health and Human Development maternal fetal medicine units network.
        Am J Obstet Gynecol. 2006; 195: 633-642
        • Crowther C.A.
        • Haslam R.R.
        • Hiller J.E.
        • Doyle L.W.
        • Robinson J.S.
        • Australian Collaborative Trial of Repeat Doses of Steroids (ACTORDS) Study Group
        Neonatal respiratory distress after repeat exposure to antenatal corticosteroids: a randomized controlled trail.
        Lancet. 2006; 367: 1913-1919
        • French N.P.
        • Hagan R.
        • Evans S.F.
        • Godfrey M.
        • Newnham J.P.
        Repeated antenatal corticosteroids: size at birth and subsequent development.
        Am J Obstet Gynecol. 1999; 180: 114-121
        • Thorp J.A.
        • Jones P.G.
        • Knox E.
        • Clark R.H.
        Does antenatal corticosteroid therapy affect birth weight and head circumference?.
        Obstet Gynecol. 2002; 99: 101-108
        • Crowther C.A.
        • Doyle L.W.
        • Haslam R.R.
        • et al.
        Outcomes at 2 years after repeat doses of antenatal corticosteroids.
        N Engl J Med. 2007; 357: 1179-1189
        • National Institutes of Health Consensus Development Panel
        Antenatal corticosteroids revisited: repeat courses; National Institutes of Health consensus statement. 2000; 17: 1-18
      2. ACOG committee opinion no. 402: antenatal corticosteroid therapy for fetal maturation.
        Obstet Gynecol. 2008; 11: 805-807
        • Liang K.Y.
        • Zeger S.L.
        Longitudinal data analysis using generalized linear models.
        Biometrika. 1986; 73: 13-22
        • Vermillion S.T.
        • Bland M.L.
        • Soper D.E.
        Effectiveness of a rescue dose of antenatal betamethasone after an initial single dose.
        Am J Obstet Gynecol. 2001; 185: 1086-1089
        • Peltoniemi O.M.
        • Kari M.A.
        • Tammela O.
        • et al.
        Randomized trial of a single repeat dose of prenatal betamethasone treatment in imminent preterm birth.
        Pediatrics. 2007; 119: 290-298
        • Vermillion S.T.
        • Soper D.E.
        • Chasedunn-Roark J.
        Neonatal sepsis after betamethasone to patients with preterm premature rupture of membranes.
        Am J Obstet Gynecol. 1999; 181: 320-327
        • Lee M.
        • Davies J.
        • Guinn D.
        • et al.
        Single vs weekly course of antenatal corticosteroids in preterm premature rupture of membranes.
        Obstet Gynecol. 2004; 103: 274-281
        • Wapner R.J.
        • Sorokin Y.
        • Mele L.
        • et al.
        Long term outcomes after repeat doses of antenatal corticosteroids.
        N Engl J Med. 2007; 357: 1190-1198

      Linked Article

      • To rescue or not to rescue: that is the question
        American Journal of Obstetrics & GynecologyVol. 200Issue 3
        • Preview
          Following the groundbreaking studies in the early 1970s which clearly demonstrated the clinical benefit of maternally-administered corticosteroids to the preterm neonate, one would have anticipated that this intervention would rapidly become a routine part of the management of preterm labor. This was not the case, however. For almost a decade following this research, routine implementation of this valuable therapy was delayed by a number of factors: varied interpretations of the efficacy, concerns about short-term safety (eg, maternal or fetal infection), fear of long-term detrimental effects, questionable utility in women with maternal diabetes or hypertensive disease, lack of dissemination of the evidence in a clinically meaningful form, innate conservatism in obstetrical practice, and a medical system that was newly evidence-based.
        • Full-Text
        • PDF
      • Correction: March 2009 (vol. 200, no. 3, page 248)
        American Journal of Obstetrics & GynecologyVol. 201Issue 4
        • Preview
          Acknowledgments for Garite TJ, Kurtzman J, Maurel K, et al. Impact of a “rescue course” of antenatal corticosteroids: a multicenter randomized placebo-controlled trial. Am J Obstet Gynecol 2009;200:248.e1-9.
        • Full-Text
        • PDF