COMPLICATIONS OF EARLY STEROID THERAPY IN A RANDOMIZED CONTROLLED TRIAL

Background: Extremely low birth weight (ELBW) infants frequently develop chronic lung disease (CLD). Meta-analyses suggest that early postnatal steroid administration may reduce this risk. Association of early cortisol insufficiency with subsequent CLD supports use of a lower stress steroid dose that may reduce complications associated with higher doses. We initiated a randomized, multicenter, controlled trial to determine whether a tapering course of stress dose steroid started on the first day would reduce the risk of CLD or death in ELBW infants. Methods: Infants with birth weight 501-1000 gm mechanically ventilated before 12 hours were eligible for the study. Infants with birth weight [is greater than] 750 gm also needed to receive Fi[O.sub.2] [is greater than or equal to] 0.30 and surfactant. Infants were randomized to receive dexamethasone (initial dose 0.15 mg/kg/d for 3 days, then tapered over 7 days) or placebo. In a factorial design, infants were also randomized to routine ventilator management or a strategy of minimal ventilator support to reduce mechanical lung injury. Results: After enrolling 220 patients (sample size estimate was 1200), the trial was halted for unanticipated adverse events. Steroid (n = 111) and placebo (n = 109) infants had similar clinical characteristics at randomization. Study Death or [O.sub.2] at PIE Steroid Drug CLD % 28 d % Rx % RX % Steroid 63 78 9 34 Placebo 69 94(*) 23(*) 51(*) Study Systolic Insulin GI Drug BP [is greater than] 80 RX % Perf % mmHg % Steroid 27 23 13 Placebo 4(*) 12(*) 3(*) (*) p = 0.05. The relative risk for the primary outcome, death or CLD at 36 postmenstrual weeks, was 0.92 (95% CI = 0.76-1.11). Fewer infants in the steroid group had pulmonary interstitial emphysema (PIE), required oxygen at 28 days, or had subsequent steroid treatment. Rates of severe intraventricular hemorrhage, periventricular leukomalacia, retinopathy of prematurity, and nosocomial infection were similar. Hypertension and hyperglycemia (insulin Rx) were more frequent in the steroid group. During the first 14 days, 14 (13%) infants in the steroid group and 3 (3%) infants in the placebo group had spontaneous gastrointestinal perforation without necrotizing enterocolitis. Perforation was associated with indomethacin exposure (p = 0.005) and there was an interaction between indomethacin and steroid (p = 0.04). No interaction was observed with ventilator strategy. Conclusion: The increased risk of gastrointestinal perforation, as well as other previously identified complications, should be considered in decisions regarding steroid treatment in ELBW infants. A R Stark, MD, FAAP, W Carlo, MD, FAAP, C Bauer, MD, FAAP E Donovan, MD, FAAP, W Oh, MD AAP, L Papile, MD, FAAP, S Shankaran, MD, FAAP, JE Tyson, MD, LL Wright, MD, FAAP, M Temprosa, K Poole, Ph, D.; for the NICHD Neonatal Research Network.3
A R Stark, MD, FAAP, W Carlo, MD, FAAP, C Bauer, MD, FAAP E Donovan, MD, FAAP, W Oh, MD AAP, L Papile, MD, FAAP, S Shankaran, MD, FAAP, JE [...]