Abstract
Objective. We designed and conducted a randomized, double-masked, controlled multicenter study to determine whether inhaled nitric oxide (INO) in term and near-term infants with congenital diaphragmatic hernia (CDH) would reduce the occurrence of death and/or the initiation of extracorporeal membrane oxygenation (ECMO).
Patients and Methods. Infants of 34 weeks gestation or more, <14 days of age with CDH, without known structural heart disease, requiring assisted ventilation for hypoxemic respiratory failure with two oxygenation indices (OIs) of 25 or more at least 15 minutes apart, were eligible for this trial. Infants were centrally randomized and then received masked treatment with 20 ppm NO or 100% oxygen as control. Infants with less than a full response to 20 ppm NO (increase in Pao2 >20 Torr) after 30 minutes were evaluated at 80 ppm NO/control study gas.
Results. The 28 control and 25 treated infants enrolled by the 13 participating centers were not significantly different at randomization for any of the measured variables including prerandomization therapies and initial OIs (45.8 ± 16.3 for controls, 44.5 ± 14.5 for INO). Death at <120 days of age or the need for ECMO occurred in 82% of control infants compared with 96% of INO infants (ns). Death occurred in 43% of controls and 48% of the INO group (ns), and ECMO treatment was used for 54% of control and 80% of INO-treated infants. There was no significant improvement in Pao2 (Δ Pao2 7.8 ± 19.8 vs 1.1 ± 7.6 Torr, ns) nor significant reduction in OI (-2.7 ± 23.4 vs 4.0 ± 14.8, ns) associated with INO treatment. Mean peak nitrogen dioxide (NO2) concentration was 1.9 ± 1.3 ppm and the mean peak methemoglobin was 1.6 ± 0.8 mg/dL. No infant had study gas discontinued for toxicity. There were no differences between the control and INO groups for the occurrence of intracranial hemorrhage, specific grades of intracranial hemorrhage, periventricular leukomalacia, brain infarction, and pulmonary or gastrointestinal hemorrhages.
Conclusions. Although the immediate short-term improvements in oxygenation seen in some treated infants may be of benefit in stabilizing responding infants for transport and initiation of ECMO, we conclude that for term and near-term infants with CDH and hypoxemic respiratory failure unresponsive to conventional therapy, inhaled NO therapy as used in this trial did not reduce the need for ECMO or death.
Congenital diaphragmatic hernia (CDH) is a malformation that occurs in approximately 1 in every 3000 to 4000 deliveries.1 The overall mortality for fetuses with isolated, potentially correctable CDH diagnosed before 24 weeks gestation is approximately 58%.2 Despite the very aggressive support required to maintain adequate gas exchange in infants with CDH who present with early-onset severe respiratory distress, there is a high rate of failure of conventional management. The major underlying pathophysiology in such infants appears to be a combination of lung hypoplasia and immaturity and persistent pulmonary hypertension, which may be further aggravated by left ventricular underdevelopment.3,4 Management of infants with CDH has included therapy directed toward the treatment of persistent pulmonary hypertension: neuromuscular blockade, sedation, alkalosis (respiratory and/or metabolic), and the use of alternative forms of ventilatory support including high-frequency oscillatory ventilation (HFOV). As of July 1995, over 2000 infants with CDH have been treated with extracorporeal membrane oxygenation (ECMO), 58% of whom survived.5 In the most recent prospective evaluation of infants with CDH, the ECMO trial in the United Kingdom, all 17 control infants with CDH died, compared with 14 deaths among the 18 ECMO-allocated infants.6 In addition to the high inherent mortality, CDH ranks among the most costly of correctable conditions, with an estimated cost per new case of $250 000, and an overall estimated yearly cost of $364 000 000 in the United States.7
The most common indication for ECMO in infants with CDH is persistent hypoxemia, thought to be secondary to persistent pulmonary hypertension. Regulation of vascular smooth muscle tone is significantly influenced by nitric oxide (NO), which is felt to be identical to the previously described endothelium-derived relaxing factor (EDRF)8-14. NO is generated enzymatically by nitric oxide synthase from the precursor L-arginine.15NO diffuses from the vascular endothelium into the vascular smooth muscle where it activates guanylate cyclase leading to the production of cyclic guanosine monophosphate.16,17 The subsequent relaxation of vascular smooth muscle by cyclic guanosine monophosphate may involve the inhibition of activation-induced elevation in cytosolic calcium concentration.18
Inhaled nitric oxide (INO) is a selective pulmonary vasodilator in animal models19-22 and in adults, improving oxygenation without producing a decrease in systemic vascular resistance.23-25 Preliminary reports26-28demonstrated that INO improved oxygenation in infants with persistent pulmonary hypertension of the newborn (PPHN) and hypoxic respiratory failure with Roberts et al using 80 ppm, whereas Kinsella27used 20 ppm followed by 6 ppm. A subsequent study found that among responsive infants there did not appear to be significant differences in the responses observed using doses from 5 to 80 ppm.29
Preliminary experience with the use of INO in infants with CDH has suggested that the majority of infants with hypoxemic respiratory failure treated shortly after delivery did not show sustained beneficial responses, but some infants showed an improvement of oxygenation after a course of ECMO.30-34
In view of these observations, a prospective multicenter randomized controlled trial was conducted to evaluate the ability of INO to prevent death or the initiation of ECMO in term and near-term infants with hypoxic respiratory failure unresponsive to aggressive conventional therapy.35 Infants with CDH and hypoxic respiratory failure were enrolled in a separate parallel study; the results of that trial are reported in this article.
MATERIALS AND METHODS
Hypotheses
The primary hypothesis of the main trial35 and the CDH trial was that the administration of INO to infants ≥34 weeks and an oxygenation index (OI) of >25 would reduce the risk of death by day 120 or discharge home (which ever came first) or the initiation of ECMO from 50% in the control group to 30% in the INO group, a relative reduction of 40%. The secondary hypotheses for the main trial were that administration of INO would lead to: an increase in Pao2, a decrease in OI and A-aDO2measured 30 minutes after initial administration of INO, a decrease in hospital days, no increase in days of assisted ventilation, incidence of air leak, bronchopulmonary dysplasia (BPD), or neurodevelopmental disability at 18 to 24 months. Infants with CDH were enrolled in a separate parallel study that enrolled patients concurrently with the main trial.
Patient Population
Any infant ≥34 weeks by best obstetric estimate who required assisted ventilation for hypoxemic respiratory failure secondary to CDH and had two OIs [(OI = (MAP × Fio2 × 100)/Pao2] ≥25 at least 15 minutes apart was eligible for participation in the trial. In addition, all infants were required to have an in-dwelling arterial line and parental permission before randomization. All centers attempted to obtain a cranial and cardiac ultrasound before entering the infant in the study. Each study center obtained institutional review board approval before enrolling infants.
Exclusion Criteria
Infants were ineligible if they were >14 days of age; had known congenital heart disease; were enrolled in conflicting clinical trials; or if a decision had been made not to provide full treatment.
Patient Management
The study protocol provided for maximal conventional treatment before randomization but it did not specify pre-enrollment management. Each participating center was required to develop a standard management strategy to be used for the duration of this trial. General management guidelines were agreed to by all centers. These included: maintenance of mean arterial blood pressure >45 mm Hg with volume infusions and/or the use of vasopressors; attempted induction of alkalosis with either hyperventilation, infusion of alkali, or both (target pH range of 7.45–7.60); and rescue treatment with a bovine surfactant (BLES, BLES Biochemicals, London, Ontario, or Survanta, Abbott Laboratories, Columbus, OH) before randomization. The mode of ventilation (conventional vs high frequency) could not be changed after randomization, except as part of weaning from assisted ventilation. The use of such therapies as sedation, analgesia, neuromuscular blockade, tolazoline, bronchodilators, and postnatal steroids was permitted. ECMO, either veno-venous or veno-arterial, was initiated when center-specific criteria were fulfilled; minimal ECMO criteria included the following:
OI >40 on two arterial blood gases (ABGs) separated by at least 30 minutes or OI >35 for 4 hours;
A-aDO2 >630 for 4 continuous hours or A-aDO2 >620 for 12 continuous hours; or
Acute deterioration/unresponsiveness to medical therapy (any 2 of 4):
Pao2 <55 Torr for >2 hours
pH <7.15, or <7.40 if alkalosis attempted, for >2 hours
Mean arterial blood pressure <40 Torr for >2 hours
Severe barotrauma (4 of 7 criteria):
1. pulmonary interstitial emphysema/pseudocyst,
2. pneumothorax/pneumomediastinum,
3. pneumoperitoneum,
4. pneumopericardium,
5. subcutaneous emphysema,
6. persistent air leak >24 hours, or
7. mean airway pressure >15 cm H2O.
Randomization
Randomization was accomplished as soon as possible after meeting eligibility criteria and obtaining a second OI ≥25. Infants were stratified by center and randomized using a permuted block design developed and managed by the George Washington University Biostatistical Coordinating Center. This system used a dedicated telephone system that included a procedure for validation and recall verification.
Study Gas Administration and Monitoring
Infants were randomized to a control group or to an INO treatment group (INO group). Control infants received 100% oxygen. If study gas could not be started within 15 minutes of the second or qualifying OI, a third ABG was obtained before study gas initiation. The third ABG was then considered the baseline ABG with respect to evaluating the response to study gas, which was started regardless of the calculated OI. Primary grade nitric oxide was supplied as 800 ppm in balanced nitrogen (Canadian Liquid Air, Montreal, Quebec or Ohmeda Inc, Liberty Corner, NJ) and was certified to be ± 1% of the analyzed component (NO), and to contain <5 ppm nitrogen dioxide (NO2). Single-stage stainless steel diffusion-free regulators were used, which were flushed to ensure that any air or other by-products such as NO2 were removed. The source gas was connected at a regulated pressure of 50 psi using Teflon tubing to the input port of a suitable flow meter, and then injected at the desired flow rate into the inspiratory circuit of a neonatal ventilator (gas flow of approximately 12 L/min or more). Quality assurance procedures were developed to insure accurate calibration of the NO/NO2 analyzers and to prevent contamination of the NO source gas.36,37
The resulting gas mixture was sampled between the injection site in the inspiratory circuit and the infant, and continuously analyzed for NO, NO2 and total oxides of nitrogen using a chemiluminescence analyzer (Model 42H, Thermo Environmental Instruments Inc, Franklin, MA, or EcoPhysics, Durnten, Switzerland) or an electrochemical analyzer, (Pulmonox II, Tofield, Alberta, Canada). Exhaled gas and exhaust from the analyzers were scavenged.
Infants were managed by the clinical team except during initiation or change of study gas concentration. Administration of study gas was masked by using designated, unmasked individuals (respiratory therapists, research nurses or physicians) in each collaborating center to obtain the randomization assignment, to set up the inhalational apparatus and the NO monitoring equipment, to adjust study gas concentrations, and to make mock adjustments to control infants. These individuals recorded the inspired oxygen, the study gas concentration, and the levels of NO and NO2 every 2 hours and after changes in ventilator settings to ensure that the appropriate study gas concentration was being administered and that NO and NO2concentrations were not increased. The analyzer readings were covered at all times to ensure masking from the clinical team. In addition, the identity of the study gas tanks were masked. During administration of study gas, the inspiratory oxygen was determined using an online oxygen sensor in the inspiratory circuit before the site of study gas administration.
For this study, a positive response was defined as an increase in arterial Pao2 above baseline 30 minutes after initial exposure to the study gas (full response >20 Torr; partial 10 to 20 Torr; no response <10 Torr). These values could be on preductal or postductal ABGs, but the comparison required sampling from the same site.
Infants were treated with the lowest study gas concentration to which they were responsive. Study gas was initiated at 20 ppm NO/control; it was continued in infants who achieved a full response 30 minutes after initiation of study gas. In infants who had less than a full response, study gas was stopped for 15 minutes if tolerated, another ABG was obtained, study gas was increased to the maximal concentration of 80 ppm NO/control, and a follow-up ABG obtained 30 minutes later. This methodology was to allow for a further assessment of the need for 80 ppm as initially reported by Roberts et al.26 Infants who had a full response to the maximal concentration remained on this increased concentration; if the response was partial, infants were continued at the lowest study gas concentration to which they had a partial response. Study gas was discontinued if they did not respond at either concentration. Study gas was also discontinued in an infant who deteriorated before the end of the initial 30-minute study gas administration period (absolute decrease in oxygen saturation >10%), and the infant was classified as a nonresponder. Infants who did not respond to the initial administration of study gas could be retried up to three times at 6-hour intervals. Crossover was not allowed between treatment groups.
Study Gas Weaning
After the initial study gas dosing, which was specified by the protocol, study gas management was at the discretion of the centers, using a recommended protocol for weaning and escalation of study gas. The maximal total duration of study gas administration was 336 hours (14 days). Weaning of study gas was only attempted if the Pao2 was more than the acceptable baseline established by each participating center (the minimal criteria being an oxygen saturation >92% and/or a Pao2 >50 Torr).
Study Gas Escalation
If, during continuous study gas administration, a deterioration occurred resulting in two OIs >25 and at least 50% more than the baseline OI measured at the last weaning attempt, the study gas concentration was doubled to a maximum of 80 ppm NO/control until a full response was obtained. The gas was returned to the pre-escalation concentration in unresponsive infants or to the lowest study gas dose to which the infant had had a partial response.
Study Gas Reinitiation
Study gas could be reinitiated after a successful wean if the patient had an OI ≥15 on two consecutive ABGs at least 30 minutes apart and had a less-than-the-maximum cumulative study gas exposure. Study gas was reinitiated at the concentration at which study gas was discontinued.
Safety Monitoring
Blood methemoglobin concentrations were measured at 1, 3, 6, and 12 hours after initiation of study gas and subsequently every 12 hours until 24 hours after gas discontinuation. Inhaled NO2concentrations were monitored continuously. Methemoglobin levels of 5 to 10% were managed with an immediate decrease in study gas concentration by 50% until the level fell to <5%. Study gas was immediately discontinued for methemoglobin level >10%. If NO2 concentrations exceeded 7 ppm, study gas was immediately discontinued; it was decreased by 50% for NO2of 5 to 7 ppm. Infants weaned off study gas for elevated methemoglobin or NO2 levels were not considered successfully weaned.
Infants were monitored for signs of increased bleeding, (ie pulmonary hemorrhage, gastrointestinal bleeding, or oozing from venipuncture sites). Cranial ultrasonography was performed whenever possible before randomization and 24 hours after final discontinuation of study gas. All readings were by local ultrasonographers and classification was based on the Papile classification.38
Statistical Considerations
Based upon the Extracorporeal Life Support Organization (ELSO) registry, we estimated that the potential population for the CDH study would be about 20% of the potential population for the main trial, which was estimated to require 125 non-CDH patients in each arm of the study to demonstrate a 40% reduction in the occurrence of death or the initiation of ECMO from 50% to 30%. Therefore, we believed that it was unlikely that we would be able to adequately test the primary and secondary hypotheses with the number of infants who could be enrolled (about 50 over 2 years). Thus, this trial was conducted to determine whether a larger trial would be indicated on the basis of the results. Tests of significance are based on t tests for means, the Wilcoxon statistic for medians, and on χ2 statistics for discrete variables. The primary analysis used the intent-to-treat paradigm.
The main and CDH pilot trials were monitored by an independent Data Safety and Monitoring Committee (DSMC). The DSMC planned two evaluations after approximately one- and two-thirds enrollment for the primary trial. It was agreed that the CDH arm would end when the primary trial was completed, unless a specific recommendation to continue enrollment was made by the DSMC.
RESULTS
The primary non-CDH study was terminated after an evaluation by the DSMC determined that INO significantly reduced the incidence of the primary outcome, death before discharge or 120 days or the initiation of ECMO in term and near-term infants with hypoxic respiratory failure. Recruitment ceased on May 2, 1996. The CDH parallel trial was terminated at the same time as the main trial, on the recommendation of the DSMC, because of a lack of observable benefit, and the very low likelihood of such an effect with continued enrollment.
Baseline
Fifty-three infants with CDH were enrolled in the trial; Table1 presents the descriptive information for this population. No significant differences between the control and INO-treated groups were noted for any of these variables. Overall, 76% of infants for whom an echocardiograph was performed (51 of 53) had evidence of PPHN, defined as either tricuspid regurgitation and/or bidirectional or right to left shunting at either the duct or foramen ovale, with similar proportions for control and INO infants. No differences between control and INO groups were noted in therapies used before randomization (Table 2). In addition, a similar number of infants in the control and INO groups had air leaks (29 vs 28%) and pulmonary hemorrhage (11 vs 4%) before randomization. Finally, there were no differences between the control and treated groups in age at randomization or initial arterial blood gas values, including Paco2, Pao2, pH, OI, or MAP (Table 3). The mean and median intervals between the time of randomization and the initiation of study gas (24.9 ± 35.5 minutes for the control group vs 51.8 ± 147.3 for the INO group, median interval 11.0 vs 12.5 minutes) were not significantly different.
Neonatal Demographics of Infants With CDH and Hypoxic Respiratory Failure
Therapies Before Randomization
Randomization Information
Outcome
Twenty-three of 28 control infants (82%) compared with 24 of 25 (96%) INO-treated infants met the primary outcome of death and/or the initiation of ECMO, results that were not significantly different (13.9% difference, 95% confidence interval −31%, 3.2%, Table4). Twelve of 28 (43%) of control infants compared with 12 of 25 (48%) of treated infants died (ns), with significantly fewer (15 of 28) control infants (54%) compared with 20 of 25 treated infants, (80%) receiving ECMO (P = .043, Table4). The age of the initiation of ECMO and the time between randomization and initiation of ECMO were not significantly different between the groups. Fifteen infants were randomized in non-ECMO centers, of whom 13 died or required ECMO (86.7%). Seven infants were transferred for ECMO (4 controls, 3 INO), 5 of whom received ECMO, and all 7 infants survived and were discharged home, whereas all 8 nontransported infants died (5 control, 3 INO). Veno-venous ECMO was used in 6 control infants and 10 NO treated infants, and 1 child in each group required conversion to veno-arterial from veno-venous ECMO. There were fewer deaths among control infants who received ECMO (4/15, 26.7%) than for INO infants who received ECMO (8/20, 40% ns). The overall mortality was 46%, 35% for infants receiving ECMO, and 67% for the remaining infants. The most frequent causes of death were withdrawal of support (8 control and 4 INO infants) and unresponsive respiratory failure [4 control and 6 INO infants (ns)]. Five infants in this trial were randomized after surgical repair (4 control and 1 INO infant). Of these 5, 1 control infant survived without ECMO and 3 of 4 control infants required ECMO and died. The single INO infant survived after ECMO.
Primary Outcome
There were no significant differences for any of the secondary outcomes between the control and treated infants (Table 5). Measurements performed 30 minutes after initiation of study gas demonstrated no significant increase in Pao2 or decreases in OI or A-aDO2 (Table 5). Twelve of 25 INO-treated infants responded (8 partial, 4 full) to 20 ppm NO/control compared with 5 of 27 controls (all partial responses,P = .024). Of all INO-treated infants who had a partial or no response at 20 ppm NO/control, only 2 infants had a partial response at 80 ppm NO/control, and no infant had a full response at 80 ppm NO/control. All control infants with no or a partial response at 20 ppm NO/control had no response to 80 ppm NO/control. The median duration of study gas administered was 1 hour for controls compared with 5 hours for INO infants (P = .003). A retrial of study gas was administered to 1 infant from each study group. Only 1 infant received study gas (NO) during transport and that infant survived. There were 4 infants evaluated at the study centers during the trial who were eligible for the trial but were not enrolled. All met ECMO criteria and received ECMO. Three died and 1 survived.
Secondary Outcomes
Adverse Events
There were no differences postrandomization between the groups in the incidence of ICH (4 INO vs 4 control). One treated infant and 2 control infants had Grade IV ICH on posttreatment cranial ultrasound. There were no significant differences in the occurrence of brain infarction (1 control vs 3 INO) or periventricular leukomalacia between the groups (2 treated vs 1 control). There were no significant differences for the occurrence of pulmonary hemorrhage, generalized oozing from venipuncture sites, or gastrointestinal bleeding between the groups. No infant required discontinuation of study gas because of toxicity secondary to elevated methemoglobin or NO2concentrations.
DISCUSSION
In this trial, we were unable to demonstrate a beneficial effect for INO in infants with CDH and hypoxic respiratory failure unresponsive to aggressive conventional therapy. While there was no difference in the occurrence of the primary outcome, significantly more INO-treated infants received ECMO. The outcome for infants who received ECMOcompares favorably with the most recent results from the ELSO Registry,3 which is a compilation of results obtained over a period of >10 years. The indices of gas exchange were improved in some (56%) infants receiving INO, but this effect was transitory and consistent with the observations of Shah et al.31 Only 4 treated infants (16%) had a 20 Torr or more increase in Pao2 with INO treatment, whereas no control infant had such an increase (P = .024).
The initial OI of both control and treated infants was well over 40, 45.8 vs 44.5, and at the second ABG, the OIs had increased to 61.6 in the control and 47.6 in the treated infants; overall, 82% of controls and 96% of the treated infants received ECMO and/or died. While we had estimated that OIs ≥25 would predict that 50% of the infants would require ECMO and/or die, the infants were more significantly compromised than we anticipated at entry, and experienced a very high rate of death or initiation of ECMO (88.7%).
Our results are consistent with the previous observations by Karamanoukian et al30 who found that early treatment of INO was not associated with significant improvements in oxygenation in infants with CDH. INO therapy post-ECMO benefited a number of infants studied by Karamoukian et al30 and Frostell et al,32 suggesting that with improvement in lung volumes and perhaps increased endogenous surfactant production, INO might exert a beneficial effect. Shah et al31 noted problems with the development of tachyphylaxis and increased plasma nitrates and nitrites.31
In the fetal lamb model of CDH, Karamanoukian et al39demonstrated that nitric oxide synthase (NOS) was present in the main pulmonary trunks of CDH lambs; however, the functional presence and activity of NO was not evaluated. North et al40 studied the Nitrofen-induced model of CDH in rats and measured both endothelial NOS (eNOS) and neuronal NOS (nNOS) in the ipsilateral CDH and control lungs. They found a similar concentration of nNOS protein in CDH vs control lungs whereas eNOS protein was decreased in the animals with CDH (58 ± 6 vs 100 ± 6%). They reported a parallel decline in eNOS mRNA in the CDH vs control lung (22 ± 8 vs 100 ± 31% in control) and suggested that the diminished eNOS gene expression may contribute to PPHN associated with CDH. These results support a possible physiologic role for INO in infants with CDH.
Karamanoukian et al41 have suggested that prophylactic surfactant therapy improved the response to INO in an animal model of CDH. In the current study, a majority of the infants received surfactant within 6 hours of randomization (64% of controls vs 80% of treated infants) and overall 82% and 84% of control and treated infants received surfactant before randomization. In the animal study of Karamanoukian et al,41 however, surfactant, to be effective, was given before the initiation of mechanical ventilation (prophylactic treatment), whereas later treatment at 30 minutes (rescue treatment) had no effect.42 The distribution of surfactant would be improved in the fluid-filled fetal lung compared with surfactant administered postnatally, especially in infants with significant lung disease requiring mechanical ventilation.
We encouraged full conventional management of these infants before the administration of INO which is an unproven therapy for such infants, and encouraged the use of surfactant, which most infants received, as well as currently accepted therapies for pulmonary hypertension. As can be seen from Table 2, the great majority of infants in the study were treated with volume support, neuromuscular blockade, sedation, the use of vasopressors, and alkalosis. Sixty-four percent of infants overall were also treated with HFOV and at randomization, more infants were receiving HFOV than conventional ventilation (55% vs 45%, ns) (Table3). It may well be that the use of these therapies delayed the initiation of the study gas and that INO used earlier may have been associated with a greater improvement. This question will need to be evaluated in an appropriately designed prospective trial.
The current study appears to be the largest prospective randomized controlled trial designed to evaluate a therapy for infants with CDH. Others include the UK ECMO trial4 and Lotze et al,43 who compared the effects of surfactant vs placebo in 17 infants with CDH who were receiving ECMO. Lotze et al found no difference in time to extubation, time on oxygen or duration of total hospitalization between the surfactant and placebo-treated groups. Nio et al44 performed a nonblinded, randomized prospective trial evaluating early (within 6 hours) and delayed (96 hours or more) surgical repair of CDH in 32 infants and reported no difference in overall survival (75% for early and 72% for delayed) or the requirement for ECMO (67% for early vs 89% for delayed) between their groups.
Bos et al45 in a nonrandomized, combined retrospective and prospective evaluation of infants with CDH noted that documented pulmonary hypertension occurred in 46% of 52 infants with CDH. They found that tolazoline did not improve oxygenation, and was associated with a significant decrease in blood pressure whereas prostacyclin did appear to improve oxygenation.
Although our numbers overall remain small, it is of interest that there were more deaths without ECMO in the control infants and more deaths with ECMO in the NO treated (both nonsignificant) suggesting the possibility that INO allowed some infants to survive to be cannulated who later died. NO therapy did not appear to unduly delay ECMO: control infants were placed on ECMO at seven hours compared with 10 hours for NO treated infants (ns), and all 7 infants who were transported for consideration of EMCO survived.
The confidence intervals of our results suggest that there may be a 3% likelihood that INO would be beneficial in reducing the occurrence of death or need for ECMO for infants with CDH and OIs similar to those seen in the current trial, balanced by the 31% possibility of a worse outcome for such infants. In addition, our trial did have sufficient power to reject a 25% reduction in the primary outcome with NO therapy using the methodology suggested by Detsky and Sackett.46Therefore, our results imply that treatment with INO, as used in this protocol, is unlikely to be of significant benefit in term and near-term infants with CDH who present with hypoxic respiratory failure and OIs >40, and may increase the need for ECMO in such infants.
However, further trials in infants with CDH may be required to assess the value of both earlier and later use of INO. In addition, although INO may help stabilize some infants during transport and/or cannulation for ECMO, its use should not delay appropriate consideration for ECMO. Our results encourage further research on the use of other supportive modalities, such as liquid ventilation,47 for critically ill infants with CDH unresponsive to conventional management.
APPENDIX: CANADIAN INHALED NITRIC OXIDE STUDY GROUP (CINOS)
Neil Finer, MD, Co-Principal Investigator**, funded by the Canadian Medical Research Council.
British Columbia Children's Hospital, Vancouver, BC–Alfonso Solimano, MD*, France Germain, RRT; Children's Hospital of Eastern Ontario, Ottawa, Ontario–Robin Walker, MD*, Anna Maria Ramirez, RRT; Foothills Hospital, Calgary, Alberta–Nalini Singhal, MD*, Leona Bourcier, RN; Health Sciences Center, Winnipeg, Manitoba–Carlos Fajardo, MD*, Valerie Cook, RN; McMaster University, Hamilton, Ontario–Haresh Kirpalani, MD*, Shelly Monkman, RRT; Montreal Children's Hospital, Montreal, Quebec–Anne Johnston, MD*, Krishna Mullahoo, RRT; Royal Alexandra Hospital, Edmonton, Alberta–Neil Finer, MD*, Abraham Peliowski, MD, Philip Etches MB, Barbara Kamstra, RN; §Texas Children's Hospital, Baylor College of Medicine, Houston, Texas–Mary Wearden, MD*, Michael Gomez, MD, Yuko Moon, MD.
NICHD NEONATAL RESEARCH NETWORK
Richard Ehrenkranz, MD, Co-Principal Investigator**, Yale University (U10HD272871), New Haven, Connecticut.
Case Western Reserve University (U10 HD21364), Cleveland, Ohio–Eileen Stork, MD*, Ellen Gorjanc, RN, Avroy A. Fanaroff, MB, BCh‡; George Washington University, The Biostatistics Center (U01 HD19897)–Joel Verter, PhD*, Naji Younes, PhD, Barbara A. Stenzel, MBA, Tonya Powers, BA; Indiana University (U10 HD27856), Indianapolis, Indiana–Greg Sokol, MD*, Diana Appel, RN, James A. Lemons, MD‡; National Institute of Child Health and Human Development, Bethesda, Maryland–Linda L. Wright, MD*‡, Sumner J. Yaffe, MD, Charlotte Catz, MD; Stanford University (U10 HD27880), Palo Alto, California–Krisa Van Meurs, MD*, William Rhine, MD*, Bethany Ball, BS, David K. Stevenson, MD‡;University of New Mexico (U10 HD27881), Albuquerque, New Mexico–Mark Crowley, MD*, Conra Backstrom, RN, Lu-Ann Papile, MD‡; Women and Infants Hospital (U10 HD27904), Providence, Rhode Island–Monica Kleinman, MD*, Angelita Hensman, RN, William Oh, MD‡.
Symbols are as follows: *, NINOS Investigator; **, NINOS Co-Principal Investigator; ‡, NICHD Neonatal Research Network Principal Investigator.
We gratefully acknowledge the valuable contributions and support provided by the respiratory therapists at each study site:
NINOS Executive Committee–Richard A. Ehrenkranz, MD, Co-Principal Investigator; Neil N. Finer, MD, Co-Principal Investigator; Anne Johnston, MD; Haresh Kirpalani, MD; Ganesh Konduri, MD; William Rhine, MD; Greg Sokol, MD; Alfonso Solimano, MD; Joel Verter, PhD; and Linda L. Wright, MD.
Writing Committee–Richard A. Ehrenkranz, MD; Neil N. Finer, MD; Joel Verter, PhD; Linda L. Wright, MD; and Naji Younes, PhD.
Data Safety and Monitoring Committee–Gordon Avery, MD, Chairman, Children's Hospital National Medical Center, Washington, DC; Mary D'Alton, MD, New England Medical Center, Boston, MA; Michael B. Bracken, PhD, Yale University, New Haven, CT; Charlotte Catz, MD, Executive Secretary, National Institute of Child Health and Human Development, Bethesda, MD; Christine A. Gleason, MD, The Johns Hopkins Hospital, Baltimore, MD; Maureen Maguire, PhD, University of Pennsylvania, Philadelphia, PA; Carol Redmond, PhD, University of Pittsburgh, Pittsburgh, PA; William Silverman, MD, Greenbrae, CA; John Sinclair, MD, McMaster University, Hamilton, Ontario; and Joel Verter, PhD (ex-offico), George Washington University, The Biostatistics Center, Rockville, MD.
ACKNOWLEDGMENTS
CINOS was supported by the Canadian Medical Research Council. NICHD Neonatal Research Centers were supported by grants (U10 HD21364, U10 HD21385, U10 HD21415, U10 HD27853, U10 HD27856, U10 HD27871, U10 HD27880, U10 HD27881, U10 HD 27904, U01 HD19897) from the National Institute Of Child Health and Human Development. Medical grade nitric oxide was provided to the NICHD Neonatal Research Network Centers by Ohmeda, a member of the BOC Group, Liberty Corner, NJ.
§Partial support for this site provided by Ross Laboratories.
Footnotes
- Received December 12, 1996.
- Accepted March 18, 1997.
This study is a collaboration of the NICHD Neonatal Research Network and the Canadian Inhaled Nitric Oxide Study Group (see Appendix).
Please address correspondence and reprint requests to: Neil M. Finer, MD, University of California, San Diego Medical Center, 200 W Arbor Dr, 8774, San Diego, CA 92103–8774.
- CDH =
- congenital diaphragmatic hernia •
- HFOV =
- high-frequency oscillatory ventilation •
- ECMO =
- extracorporeal membrane oxygenation •
- NO =
- nitric oxide •
- NO2 =
- nitrogen dioxide •
- EDRF =
- endothelium-derived relaxing factor •
- INO =
- inhaled nitric oxide •
- PPHN =
- persistent pulmonary hypertension of the newborn •
- OI =
- oxygenation index •
- BPD =
- bronchopulmonary dysplasia •
- ELSO =
- Extracorporeal Life Support Organization •
- DSMC =
- Data Safety and Monitoring Committee •
- NOS =
- nitric oxide synthase •
- eNOS =
- endothelial NOS •
- nNOS =
- neuronal NOS
REFERENCES
- Copyright © 1997 American Academy of Pediatrics
In this issue
Jump to section
Related Articles
Cited By...
- Inhaled Pulmonary Vasodilators in the Neonatal and Pediatric ICU
- The CoDiNOS trial protocol: an international randomised controlled trial of intravenous sildenafil versus inhaled nitric oxide for the treatment of pulmonary hypertension in neonates with congenital diaphragmatic hernia
- Diagnosis and management of congenital diaphragmatic hernia: a clinical practice guideline
- Congenital Diaphragmatic Hernia: Maximizing Survival
- Pediatric Pulmonary Hypertension: Guidelines From the American Heart Association and American Thoracic Society
- Inhaled Nitric Oxide Use in Neonates With Congenital Diaphragmatic Hernia
- Congenital diaphragmatic hernia
- Congenital Diaphragmatic Hernia: Updates and Outcomes
- Evidence-Based Clinical Practice Guideline: Inhaled Nitric Oxide for Neonates With Acute Hypoxic Respiratory Failure
- Neonatal extracorporeal membrane oxygenation: practice patterns and predictors of outcome in the UK
- Medical and Financial Impact of a Neonatal Extracorporeal Membrane Oxygenation Referral Center in the Nitric Oxide Era
- Congenital Diaphragmatic Hernia: Real Improvements in Survival
- Survival in Early- and Late-Term Infants With Congenital Diaphragmatic Hernia Treated With Extracorporeal Membrane Oxygenation
- The Endothelin A Receptor Antagonists PD 156707 (CI-1020) and PD 180988 (CI-1034) Reverse the Hypoxic Pulmonary Vasoconstriction in the Perinatal Lamb
- Randomized, Prospective Study of Low-Dose Versus High-Dose Inhaled Nitric Oxide in the Neonate With Hypoxic Respiratory Failure
- Use of Inhaled Nitric Oxide
- Recent advances in neonatology
- Management of congenital diaphragmatic hernia
- Effect of dexamethasone on endothelial nitric oxide synthase in experimental congenital diaphragmatic hernia
- Recent advances: care of near term infants with respiratory failure
- Inhaled nitric oxide in neonates