ORIGINAL www.jpeds.

com • THE JOURNAL OF PEDIATRICS

ARTICLES
Race Effects of Inhaled Nitric Oxide in Preterm Infants: An Individual
Participant Data Meta-Analysis
Lisa M. Askie, PhD, MPH, BN1, Lucy C. Davies, MSc, BSc1, Michael D. Schreiber, MD2, Anna Maria Hibbs, MD, MSCE3,
Philip L. Ballard, MD, PhD4, and Roberta A. Ballard, MD4

Objective To assess whether inhaled nitric oxide (iNO) improves survival without bronchopulmonary dysplasia
(BPD) for preterm African American infants.
Study design An individual participant data meta-analysis was conducted, including 3 randomized, placebo-
controlled trials that enrolled infants born at <34 weeks of gestation receiving respiratory support, had at least 15%
(or a minimum of 10 infants in each trial arm) of African American race, and used a starting iNO of >5 parts per
million with the intention to treat for 7 days minimum. The primary outcome was a composite of death or BPD.
Secondary outcomes included death before discharge, postnatal steroid use, gross pulmonary air leak, pulmonary
hemorrhage, measures of respiratory support, and duration of hospital stay.
Results Compared with other races, African American infants had a significant reduction in the composite outcome
of death or BPD with iNO treatment: 49% treated vs 63% controls (relative risk, 0.77; 95% CI, 0.65-0.91; P = .003;
interaction P = .016). There were no differences between racial groups for death. There was also a significant dif-
ference between races (interaction P = .023) of iNO treatment for BPD in survivors, with the greatest effect in African
American infants (P = .005). There was no difference between racial groups in the use of postnatal steroids, pul-
monary air leak, pulmonary hemorrhage, or other measures of respiratory support.
Conclusion iNO therapy should be considered for preterm African American infants at high risk for BPD. iNO to
prevent BPD in African Americans may represent an example of a racially customized therapy for infants. (J Pediatr
2018;193:34-9).

B
ronchopulmonary dysplasia (BPD) in preterm infants (<34 weeks of gestation) remains a highly significant and costly
cause of mortality and morbidity.1 Based on preclinical studies in a variety of animal models,2-8 there is evidence that
inhaled nitric oxide (iNO) may prevent some of the developmental changes leading to BPD. A number of clinical trials
of iNO to prevent BPD in preterm infants have been conducted, with varied results.9 Schreiber et al demonstrated improved
survival without BPD.10 In this trial, intubated infants were started on 10 parts per million (ppm) of iNO for 7 days and the
enrolled cohort was 70% African American. In the Nitric Oxide to prevent Chronic Lung Disease (NOCLD) trial, infants had
a higher exposure to iNO (starting at 20 ppm, tapered over 24 days) than in other trials.11 In addition, African American and
Hispanic infants had better survival without BPD than other races (P = .055 for interaction between race and treatment).
In 2011, the Meta-analysis of Preterm Patients on Inhaled Nitric Oxide (MAPPiNO) Collaboration was formed to further
explore the variation in outcomes from the various trials by undertaking an individual participant data (IPD) meta-analysis of
12 trials meeting prespecified inclusion criteria.12 Some of the advantages of IPD meta-analysis over meta-analysis based on
aggregate data include improved uniformity in defining participant characteristics and outcome measures, including sub-
group definitions, the ability to adjust for the nonindependence of siblings within the dataset, and the opportunity to collect
information on longer term outcomes.13 The objectives of the MAPPiNO IPD meta-analysis were to determine whether iNO
treatment in preterm infants receiving assisted ventilation improved survival rates with less BPD, and whether the outcome of
trials differed according to participant- or intervention-related factors, including gestational age at birth, birth weight, multi-
plicity, race, postnatal age at the time of randomization, iNO dosage and duration of administration, as well as other comorbidities
of prematurity. In MAPPiNO, one of the prespecified subgroup analyses was a start-
ing iNO dose categorized as 5 ppm or less, or greater than 5 ppm, by trial design.

From the 1National Health and Medical Research Council

Clinical Trials Centre, University of Sydney, Sydney,
BPD Bronchopulmonary dysplasia Australia; 2University of Chicago, Chicago, IL;
3
iNO Inhaled nitric oxide Department of Pediatrics, Case Western Reserve
University, and Rainbow Babies and Children’s Hospital,
IPD Individual participant data Cleveland, OH; and 4University of California, San
MAPPiNO Meta-analysis of Preterm Patients on Inhaled Nitric Oxide Francisco, CA
NEWNO NEWborns treated with inhaled Nitric Oxide Funded by an unrestricted grant from IKARIA, INC (to
NOCLD Nitric Oxide to prevent Chronic Lung Disease R.B). All authors were members of the MAPPiNO
Collaboration. The authors declare no conflicts of interest.
ppm Parts per million
RiNOP Race and inhaled Nitric Oxide in Preterm Infants 0022-3476/$ - see front matter. © 2017 Elsevier Inc. All rights
RR Relative risk reserved.
https://doi.org10.1016/j.jpeds.2017.10.004

34

The effect of iNO treatment on the composite outcome of death
or BPD in the higher starting dose group was significantly larger
than in the lower starting dose group (relative risk [RR], 0.83
[95% CI, 0.74-0.95] vs RR, 1.00 [95% CI, 0.94-1.06], respec-
tively; P = .02 for interaction). The trials starting with an iNO
dose of greater than 5 ppm, all had an intention to treat infants
for a minimum of 7 days providing a greater overall expo-
sure to iNO for infants in those trials. A further prespecified
subgroup analysis of the MAPPiNO IPD meta-analysis was the
effect of race, which showed a trend toward greater treat-
ment effect in non-White infants compared with Whites, but
which was not statistically significant.14
The 2010 National Institutes of Health Consensus Confer-
ence on the use of iNO in the preterm infants15 concluded that,
although there was insufficient evidence to recommend the
routine use of iNO to prevent CLD, “the positive results from
one multicenter trial,11 characterized by later timing, higher
dose and longer duration require confirmation.” This recom-
mendation, in addition to the publication of the MAPPiNO
results, led to an additional randomized trial assessing the effect
of iNO in preterm infants, the NEWborns treated with inhaled
Nitric Oxide (NEWNO) trial,16 which used the longer, higher
dose regimen, similar to the NOCLD trial. Although there was
a baseline imbalance in enrollment by race (White 52% of iNO-
treated infants vs 38% of controls) in this trial, there was a trend
toward better survival without BPD in African American infants
(45% in iNO group vs 35% in placebo group; P = .213) but
not in Whites. To further analyze these observations from 3
trials,10,11,16 and the MAPPINO analysis12 related to iNO dose
and race, we performed an IPD meta-analysis, named the Race
and inhaled Nitric Oxide in Preterm Infants (RiNOP) study.
Methods
Studies were considered eligible for this IPD meta-analysis if
they had randomly assigned preterm infants born at less than
34 weeks of gestation who were receiving respiratory support,
to either iNO or a placebo. The studies also fulfilled the fol-
lowing criteria: they enrolled at least 50 infants and re-
mained blinded to treatment allocation; a minimum of 15%
of the enrolled infants or at least 10 infants in each treat-
ment arm were of African American race, selected to allow for
a sufficient number of African American infants for analysis;
the trial protocol started infants on greater than 5 ppm of iNO
with the intention to treat infants for at least 7 days; and the
investigators had IPD available to be analyzed. All of the eli-
gible individual studies had ethics committee/institutional
review board approval and parental consent. The investiga-
tors of each eligible trial were contacted and invited to join
the collaborative group and to supply line-by-line raw data for
each individual, randomly assigned participant in their trial.
These data were then checked for any missing information,
errors, or inconsistency with previously published reports. The
main outcomes were calculated for each participant to align
with the prespecified definitions as outlined in the eligible trial
protocols. All randomized participants with outcome data
Volume 193 • February 2018
available were included in the analyses, which were per-
formed on an intention-to-treat basis.
For each of the outcomes, a 1-stage approach to analysis was
undertaken so that the IPD from the eligible trials were in-
cluded in a single model. This enabled the variation across trials
to be accounted for within the model by including a fixed trial
effect. For binary outcomes, RRs and 95% CIs were obtained
using log-binomial regression models and for continuous vari-
ables mean differences and 95% CIs between treatment groups
were obtained using linear regression models. A treatment-
by-trial interaction term was tested for each outcome to assess
the heterogeneity of treatment effect across trials.17 Models as-
sessing whether race is predictive of treatment benefit in-
cluded a test for interaction between race and treatment and
were adjusted for trial. Because multiple births occurred fre-
quently (17%) in the eligible trials, correlation between out-
comes owing to multiple births was taken into account using
generalized estimating equations.18 The primary endpoint of
this IPD meta-analysis was a composite outcome of death or
BPD, defined as requiring respiratory support or supplemen-
tal oxygen at 36 weeks’ postmenstrual age. Schreiber also used
chest radiographs as part of the BPD definition, and both the
NOCLD and NEWNO trials used the oxygen reduction test
to determine the need for supplemental oxygen. Secondary end-
points included death before discharge, postnatal steroid use,
gross pulmonary air leak, pulmonary hemorrhage, measures
of respiratory support, and duration of hospital stay. Sub-
group analyses for all endpoints were performed by race/
ethnicity category (defined as either non-Hispanic African
American race, non-Hispanic white race, Hispanic ethnicity,
other race, or unknown) to evaluate the effect of maternal self-
reported race on outcome. All statistical analyses were per-
formed in SAS 9.4 (SAS Institute, Cary, North Carolina) and
were not adjusted for multiple testing.
Results
There were 3 trials that met the prespecified criteria for
inclusion in the IPD with outcome data available for 1240
preterm infants. Schreiber10 and NOCLD11 were the 2 trials in-
cluded in the original MAPPiNO IPD meta-analysis that started
with greater than 5 ppm iNO by protocol. A third trial, the
NEWNO trial,16 which was conducted subsequent to the
MAPPiNO publication, also met the criteria (Figure 1; avail-
able at www.jpeds.com) and the IPD were made available to
the data management team at the National Health and Medical
Research Council Clinical Trials Centre at the University of
Sydney by IKARIA INC, the sponsoring organization. Table I
presents the key protocol specifications of the included trials.
The NOCLD trial used iNO starting at 20 ppm for 4 days, then
decreasing to 10, 5, and 2 ppm at weekly intervals. NEWNO
used a similar schedule, but remained at 5 ppm until 24 days.
The Schreiber trial started at 10 ppm and decreased to 5 ppm
with an intention to treat for at least 7 days. Table II outlines
the key baseline characteristics of the included trials and shows
the breakdown of racial groups by treatment allocation for the
35
THE JOURNAL OF PEDIATRICS • www.jpeds.com Volume 193

Table I. Key protocol elements of the included trials

Trial names (design)
Registration number
Intervention
Control
Primary outcome
Country of recruitment
Number randomized,
inclusion criteria
Infants assigned to
intervention group, n (%)
Infants assigned to control
group, n (%)
Multiples in trial, n (%)
Schreiber
(2 × 2 factorial RCT)
Before registration required
iNO starting at 10 ppm for 12-24 hours,
then 5 ppm for 6 days, then weaned
by 1 ppm every 6 hours if PaO2 did
not decrease by more than 15% until
extubation, OR high-frequency
oscillatory ventilation OR both. iNO
discontinued after extubation.
Placebo (nitrogen)
Death or CLD (supplemental oxygen and
CXR showing persistent parenchymal
lung disease at 36 weeks PMA)
among surviving infants.
USA
207 infants <34 weeks and <2000 g
birth weight, <72 hours of age, and
intubate/ventilated for RDS, having
had exogenous surfactant.
105 (51)
102 (49)
26 (13)
NOCLD NEWNO
(RCT) (RCT)
NCT00000548 NCT00931632
iNO at 20 ppm initial dose for 48 to 96 iNO at 20 ppm for 72-96 hours,
hours, then dose subsequently decreased to 10 ppm for 1 week,
decreased to 10, 5 and 2 ppm at then to 5 ppm until 24 days. iNO
weekly intervals, with a minimum continued after extubation.
treatment duration of 24 days. iNO
continued after extubation.
Placebo (nitrogen) Placebo (nitrogen)
Survival without BPD at 36 weeks PMA Survival without BPD at 36 weeks PMA
by room air reduction test. by room air reduction test.
USA USA and Canada
582 infants <1250 g and <32 weeks on 451 infants <30 weeks and <1250 g on
assisted ventilation at 7-21 days assisted ventilation or, if <800g, CPAP
(or, if <800 g, on CPAP). at 5-14 days.
294 (51) 229 (51)
288 (49) 222 (49)
84 (14) 103 (30)

CLD, Chronic lung disease; CPAP, continuous positive airway pressure; CXR, chest radiographs; RCT, randomized controlled trial; RDS, respiratory distress syndrome.

combined data. Table III (available at www.jpeds) provides a for these high-risk ventilated preterm infants. There was no
breakdown of baseline characteristics by race/ethnicity. difference in the response to iNO for infants entered into the
Figure 2 lists the main outcomes by race/ethnicity. Com- trials of White (P = .87), Hispanic (P = .14), or other mater-
pared with infants of other races, African American infants nal self-reported races/ethnicities (P = .21). There was no dif-
had a significant reduction in the composite outcome of death ference between racial groups for death at any time, death at
or BPD with iNO treatment: 49% in the iNO treated infants discharge, or death at 36 weeks. There was, however, a signifi-
compared with 63% in the control infants (RR, 0.77; 95% CI, cant difference between races (interaction P = .023) for the effect
0.65-0.91; P = .003), with an interaction P value of .016. In the of iNO treatment on the incidence of BPD in survivors, with
placebo group, there was no difference in death or BPD by race the greatest effect being seen in African American infants
(P = .005). There was no difference between racial groups in
use of postnatal steroids or for incidence of pulmonary air leak
Table II. Baseline characteristics by trial or race/ethnicity, or pulmonary hemorrhage (data not shown).
by treatment group Table IV presents outcomes for all infants by treatment using
an IPD meta-analysis of participants in the 3 trials of high iNO
Overall, n (%) or mean (SD)
exposure included in this study. Overall, the composite outcome
Characteristics iNO Placebo of death or BPD was reduced in infants treated with iNO (58%)
Male vs control infants (65%; RR, 0.90; 95% CI, 0.82-0.98; P = .018).
Schreiber 63 (60%) 56 (55%) There was no difference in death at any time point between
NOCLD 155 (53%) 162 (56%)
NEWNO 115 (50%) 116 (52%) the 2 treatment groups. Meta-analyzed data from the 3 in-
Total 333 (53%) 334 (55%) cluded trials also showed a statistically significant reduction
Gestational age at birth (wk) in BPD for surviving infants treated with iNO (RR, 0.89; 95%
Schreiber 27.4 (2.5) 27.0 (2.8)
NOCLD 25.7 (1.4) 25.7 (1.5) CI, 0.80-0.98; P = .02). None of the prespecified continuous
NEWNO 25.6 (1.4) 25.6 (1.5) outcomes, including measures of respiratory support and hos-
Total 26.0 (1.8) 25.9 (1.9) pital stay, showed significant differences for iNO-treated com-
Birth weight (g)
Schreiber 1017 (369) 949 (387) pared with placebo infants, nor were there any differences by
NOCLD 766 (161) 759 (155) race (Tables V and VI; available at www.jpeds.com).
NEWNO 724 (160) 750 (163)
Total 792 (234) 787 (226)
Race
White, not Hispanic 308 (56%) 242 (44%)
Discussion
Hispanic 87 (45%) 105 (55%)
African American 202 (47%) 226 (53%) The results of this IPD meta-analysis revealed that there was
Other 31 (44%) 39 (56%) no respiratory benefit at 36 weeks’ postmenstrual age for infants
Total 628 (51%) 612 (49%)
whose maternal self-reported race was White, Hispanic, or
36 Askie et al
February 2018 ORIGINAL ARTICLES

Figure 2. Main outcomes by race/ethnicity.

classified as other and who were treated with iNO starting at number needed to treat of 7. Between racial groups there
a dose of greater than 5 ppm. However, treated African Ameri- were no imbalances in other factors that potentially influ-
can infants in the 3 trials included in this meta-analysis had enced the death or BPD outcome (Table III). We did not see
a statistically significant benefit from iNO treatment, with a a difference in duration of hospitalization or resource use in

Table IV. Main outcomes by trial

Overall n (%) Adjusted* 95% CI
Outcomes Trial iNO Placebo RR LCL UCL P value Interaction P value
Death or BPD at 36 weeks PMA Schreiber 51 (49) 64 (63) 0.77 0.60 0.99 .042
NOCLD† 165 (56) 183 (64) 0.88 0.77 1.01 .069
NEWNO 146 (64) 148 (67) 0.96 0.84 1.09 .516
Heterogeneity: I2 = 31% Overall 362 (58) 395 (65) 0.90 0.82 0.98 .018 .316
Death at any time‡ Schreiber 16 (15) 23 (23) 0.68 0.38 1.22 .198
NOCLD 22 (7) 21 (7) 1.03 0.58 1.83 .929
NEWNO 19 (8) 15 (7) 1.24 0.65 2.36 .517
Heterogeneity: I2 = 0% Overall 57 (9) 59 (10) 0.93 0.66 1.30 .661 .359
Death at 36 weeks PMA Schreiber 14 (13) 20 (20) 0.70 0.37 1.33 .276
NOCLD 16 (5) 18 (6) 0.87 0.45 1.67 .678
NEWNO 19 (8) 15 (7) 1.24 0.65 2.36 .517
Heterogeneity: I2 = 0% Overall 49 (8) 53 (9) 0.89 0.62 1.29 .541 .431
BPD in survivors - trialist defined Schreiber 37 (41) 44 (54) 0.50 0.14 1.80 .286
NOCLD 149 (51) 165 (57) 0.86 0.74 1.01 .065
NEWNO 127 (61) 133 (65) 0.94 0.81 1.09 .383
Heterogeneity: I2 = 0% Overall 313 (53) 342 (60) 0.89 0.80 0.98 .020 .445

LCL, Lower confidence limit; PMA, postmenstrual age; UCL, upper confidence limit.
*Model included adjustments for trial, treatment and a treatment-by-trial interaction term.
†Assessed in infants enrolled between 7 and 14 days.
‡There are 3 infants in the Schreiber trial whose date of death is after discharge and one infant whose date of death is unknown. For the NOCLD and NEWNO trials, there is no information for
infants who died after discharge.

Race Effects of Inhaled Nitric Oxide in Preterm Infants: An Individual Participant Data Meta-Analysis 37
THE JOURNAL OF PEDIATRICS • www.jpeds.com
the meta-analysis, although there was a significant decrease
noted in the NOCLD trial.19 Of note, African American race
was not associated with fewer deaths or less BPD in the placebo
group, indicating a similar risk for untreated infants, regard-
less of race. Thus, the apparent benefit of iNO for all infants
in these trials is largely derived from a significant treatment
response in African American infants.
There are a number of reported observations supporting
racial disparity in the production, metabolism, and/or func-
tion of Nitric Oxide (NO). Black or Asian maternal race com-
pared with white race is associated independently with persistent
pulmonary hypertension of the newborn, a disease that is likely
related to endogenous NO signaling and that is responsive to
iNO therapy.20 Several studies indicate reduced bioavailability
of NO in African Americans compared with Whites, likely in
part due to increased oxidation of NO. In laboratory studies,
release of NO from umbilical venous endothelial cells has been
shown to be substantially less for African Americans com-
pared with White infants.21,22 Levels of urinary NO metabo-
lites are higher in self-reported White compared with African
American and Hispanic premature infants, both with and
without iNO treatment, perhaps reflecting differences in NO
metabolism and, thus, bioavailability.23 In adults, African Ameri-
cans have increased frequency of hypertension and cardiovas-
cular disease, and a NO-targeted medication (hydralazine
hydrochloride and isosorbide dinitrate) is an indicated therapy
for heart failure specifically in African Americans.23-25 It is likely
that these racial differences in NO bioavailability and re-
sponse are due to genetic variation in key genes on NO path-
ways, and both racial and ethnic differences exist for allele
frequency of single nucleotide polymorphisms in many related
genes, including NOS1 and NOS3.26-30
Some limitations of the RiNOP study include that the era
of enrollment for the trials was before 2011, so the results
may not reflect recent advances in infant intensive care that
could impact iNO response and respiratory outcome. In ad-
dition, there were small differences in the definition of BPD
between studies. Race was assigned by maternal self-report
and, thus, does not capture the proportion of African, Euro-
pean, and Native American ancestry within individuals.
However, we have observed a high concordance between genetic
African ancestry and self-reported African American race in
the Trial Of Late SURFactant trial cohort (unpublished data).31
One trial (Schreiber 2003)10 was a single-center study, but the
other two (NOCLD, NEWNO)11,16 were multicenter trials in-
volving a total of 54 different sites. Despite the similarity
between racial groups in gestational age (Table III), it is pos-
sible that African American infants were more developmen-
tally mature, as has been proposed based on population data
for gestational age at delivery,32 and that maturity impacts
responsiveness to iNO. However, this was not observed in our
analysis by gestational age (data not presented). Although the
3 trials included a reasonable number of African American
infants, there were relatively few infants of other racial back-
grounds and thus there was insufficient power to examine
the efficacy of iNO in other racial groups. However, it is of
particular interest that some Hispanic populations have a
38
Volume 193
substantial proportion of African ancestry, which may also
convey iNO responsiveness.33
Another limitation of this study is the lack of longer term
outcome data to analyze via IPD meta- analysis at 40 weeks’
postmenstrual age or later in infancy. Although both the Sch-
reiber and NOCLD trials reported improved longer term
outcomes,34,35 there was no difference in either pulmonary or
neurodevelopmental outcomes in treated vs control infants in
the NEWNO study.16 It is well-known that many other envi-
ronmental and social factors influence the long-term outcome
in these high-risk infants. A multicenter randomized trial
targeting this population would provide further valuable
information.
Our IPD meta-analysis of 3 randomized, controlled trials
of iNO supports further investigation of genetic ancestry and
the potential contribution of genetic variants related to NO
and biological pathways leading to BPD. Because iNO is safe
and well-tolerated by infants, and may reduce later respira-
tory morbidity34 as well as BPD/death with a number needed
to treat of 7, we suggest that iNO therapy should be consid-
ered for preterm African American infants at high risk for BPD.
There was also a 9 percentage point reduction in the rate of
BPD in survivors (46% iNO vs 55% placebo) in Hispanic
infants (many of whom have some African ancestry)31 that, al-
though not statistically significant, may provide rationale for
also considering therapy in this population. We conclude that
iNO to prevent BPD in African Americans represents an
example of a racially customized therapy for infants. ■
Submitted for publication May 31, 2017; last revision received Aug 12, 2017;
accepted Oct 6, 2017
References
1. Stoll BJ, Hansen NI, Bell EF, Walsh MC, Carlo WA, Shankaran S, et al.
Trends in care practices, morbidity, and mortality of extremely preterm
neonates, 1993-2012. JAMA 2015;314:1039-51.
2. Balasubramaniam V, Maxey AM, Morgan DB, Markham NE, Abman SH.
Inhaled NO restores lung structure in eNOS-deficient mice recovering
from neonatal hypoxia. Am J Physiol Lung Cell Mol Physiol 2006;291:L119-
27.
3. Ballard PL, Gonzales LW, Godinez RI, Godinez MH, Savani RC, McCurnin
DC, et al. Surfactant composition and function in a primate model of
infant chronic lung disease: effects of inhaled nitric oxide. Pediatr Res
2006;59:157-62.
4. Bland RD, Albertine KH, Carlton DP, MacRitchie AJ. Inhaled nitric oxide
effects on lung structure and function in chronically ventilated preterm
lambs. Am J Respir Crit Care Med 2005;172:899-906.
5. Martin RJ, Mhanna MJ, Haxhiu MA. The role of endogenous and
exogenous nitric oxide on airway function. Semin Perinatol 2002;26:432-
8.
6. McCurnin DC, Pierce RA, Chang LY, Gibson LL, Osborne-Lawrence S,
Yoder BA, et al. Inhaled NO improves early pulmonary function and modi-
fies lung growth and elastin deposition in a baboon model of neonatal
chronic lung disease. Am J Physiol Lung Cell Mol Physiol 2005;288:L450-
9.
7. Shaul PW, Afshar S, Gibson LL, Sherman TS, Kerecman JD, Grubb PH,
et al. Developmental changes in nitric oxide synthase isoform expres-
sion and nitric oxide production in fetal baboon lung. Am J Physiol Lung
Cell Mol Physiol 2002;283:L1192-9.
8. Tang J-R, Markham NE, Lin Y-J, McMurtry IF, Maxey A, Kinsella JP, et al.
Inhaled nitric oxide attenuates pulmonary hypertension and improves lung
Askie et al
February 2018
growth in infant rats after neonatal treatment with a VEGF receptor in-
hibitor. Am J Physiol Lung Cell Mol Physiol 2004;287:L344-51.
9. Barrington KJ, Finer N, Pennaforte T. Inhaled nitric oxide for respira-
tory failure in preterm infants. Cochrane Database Syst Rev 2017;(1):
CD000509.
10. Schreiber MD, Gin-Mestan K, Marks JD, Huo D, Lee G, Srisuparp P.
Inhaled nitric oxide in premature infants with the respiratory distress syn-
drome. NEJM 2003;349:2099-107.
11. Ballard RA, Truog WE, Cnaan A, Martin RJ, Ballard PL, Merrill JD, et al.
Inhaled nitric oxide in preterm infants undergoing mechanical ventila-
tion. NEJM 2006;355:343-53.
12. Askie LM, Ballard RA, Cutter G, Dani C, Elbourne D, Field D, et al. Inhaled
nitric oxide in preterm infants: a systematic review and individual patient
data meta-analysis. BMC Pediatr 2010;10:15.
13. Tierney JF, Vale C, Riley R, Smith CT, Stewart L, Clarke M, et al. Indi-
vidual participant data (IPD) meta-analyses of randomised controlled trials:
guidance on their use. PLoS Med 2015;12:e1001855.
14. Askie LM, Ballard RA, Cutter GR, Dani C, Elbourne D, Field D, et al.
Inhaled nitric oxide in preterm infants: an individual-patient data meta-
analysis of randomized trials. Pediatrics 2011;128:729-39.
15. Cole FS, Alleyne C, Barks JD, Boyle RJ, Carroll JL, Dokken D, et al. NIH
Consensus Development Conference statement: inhaled nitric-oxide
therapy for premature infants. Pediatrics 2011;127:363-9.
16. Hasan SU, Potenziano J, Konduri GG, Perez JA, Van Meurs KP, Walker
W, et al.; For the Newborns Treated With Nitric Oxide (NEWNO) Trial
Group. Effect of inhaled nitric oxide on survival without bronchopul-
monary dysplasia in preterm infants a randomized clinical trial. JAMA
Pediatr 2017;71:1081-9.
17. Higgins J, Thompson SG. Quantifying heterogeneity in a meta-analysis.
Stat Med 2002;21:1539-58.
18. Liang K, Zeger SL. Longitudinal data analysis using generalized linear
models. Biometrika 1986;73:13-22.
19. Zupancic JA, Hibbs AM, Palermo L, Truog WE, Cnaan A, Black DM, et al.
Economic evaluation of inhaled nitric oxide in preterm infants under-
going mechanical ventilation. Pediatrics 2009;124:1325-32.
20. Hernández-Díaz S, Van Marter LJ, Werler MM, Louik C, Mitchell AA. Risk
factors for persistent pulmonary hypertension of the newborn. Pediat-
rics 2007;120:e272-82.
21. Kalinowski L, Dobrucki IT, Malinski T. Race-specific differences in en-
dothelial function: predisposition of African Americans to vascular dis-
eases. Circulation 2004;109:2511-7.
22. Mason RP, Kalinowski L, Jacob RF, Jacoby AM, Malinski T.
Nebivolol reduces nitroxidative stress and restores nitric oxide
bioavailability in endothelium of black Americans. Circulation
2005;112:3795-801.
23. Ballard PL, Keller RL, Black DM, Durand DJ, Merrill JD, Eichenwald EC,
et al. Inhaled nitric oxide increases urinary nitric oxide metabolites
Race Effects of Inhaled Nitric Oxide in Preterm Infants: An Individual Participant Data Meta-Analysis
ORIGINAL ARTICLES
and cyclic guanosine monophosphate in premature infants: relation-
ship to pulmonary outcome. Am J Perinatol 2015;32:225-32.
24. Ishizawar D, Yancy C. Racial differences in heart failure therapeutics. Heart
Fail Clin 2010;6:65-74.
25. Khan BV, Rahman ST, Haque T, Merchant N, Bhaheetharan S, Harris IIIJ,
et al. Vascular effects of nebivolol added to hydrochlorothiazide in African
Americans with hypertension and echocardiographic evidence of
diastolic dysfunction: the NASAA study. J Cardiovasc Pharmacol Ther
2012;17:291-7.
26. Fujihara J, Yasuda T, Kawai Y, Morikawa N, Arakawa K, Koda Y, et al.
First survey of the three gene polymorphisms (PON1 Q192R, eNOS E298D
and eNOS C-786T) potentially associated with coronary artery spasm in
African populations and comparison with worldwide data. Cell Biochem
Funct 2011;29:156-63.
27. Liu R, Geng P, Ma M, Yu S, Wang X, Zhang W, et al. Association between
endothelial nitric oxide synthase gene polymorphism (T-786C) and
ischemic stroke susceptibility: a meta- analysis. Int J Neurosci 2014;124:642-
51.
28. Malhotra S, Poole J, Davis H, Dong Y, Pollock J, Snieder H, et al. Effects
of NOS3 Glu298Asp polymorphism on hemodynamic reactivity to stress:
influences of ethnicity and obesity. Hypertension 2004;44:866-71.
29. McNamara DM, Tam SW, Sabolinski ML, Tobelmann P, Janosko K,
Venkitachalam L, et al. Endothelial nitric oxide synthase (NOS3) poly-
morphisms in African Americans with heart failure: results from the
A-HeFT trial. J Card Fail 2009;15:191-8.
30. Yanamandra K, Napper D, Pramanik A, Bocchini JA Jr, Dhanireddy R.
Endothelial nitric oxide synthase genotypes in the etiology of retinopathy
of prematurity in premature infants. Ophthalmic Genet 2010;31:173-7.
31. Keller RL, Oh SS, Torgerson D, Ballard PL, Burchard E, Huntsman S, et al.
Maternal race/ethnicity (RE) and infant genetic ancestry associates with
survival without bronchopulmonary dysplasia (BPD) in preterm new-
borns treated with inhaled nitric oxide (iNO). J Invest Med 2016;64:263.
(abstract).
32. Robillard PY, Hulsey TC, Alexander GR, Sergent MP, de Caunes F,
Papiernik E. Hyaline membrane disease in black newborns: does fetal lung
maturation occur earlier? Eur J Obstet Gynecol Reprod Biol 1994;55:157-
61.
33. Price Alkes L, Patterson N, Yu F, Cox David R, Waliszewska A, McDonald
Gavin J, et al. A genomewide admixture map for Latino populations. Am
J Hum Genet 2007;80:1024-36.
34. Hibbs AM, Walsh MC, Martin RJ, Truog WE, Lorch SA, Alessandrini E,
et al. One-year respiratory outcomes of preterm infants enrolled in the
nitric oxide([H1]) (to prevent) chronic lung disease trial of inhaled nitric
oxide. J Pediatr 2008;153:525-9.
35. Mestan KKL, Marks JD, Hecox K, Huo D, Schreiber MD.
Neurodevelopmental outcomes of premature infants treated with inhaled
nitric oxide. N Engl J Med 2005;353:23-32.
39
THE JOURNAL OF PEDIATRICS • www.jpeds.com Volume 193

Table III. Baseline characteristics by race/ethnicity and

treatment
Overall, n (%) or mean (SD)
Outcomes iNO Placebo
Male
White, not Hispanic 156 (51%) 142 (59%)
Hispanic 50 (57%) 55 (52%)
African American 110 (54%) 115 (51%)
Other 17 (55%) 22 (56%)
Overall 333 (53%) 334 (55%)
Multiples
White, not Hispanic 77 (25.0%) 52 (21.5%)
Hispanic 11 (12.6%) 14 (13.3%)
African American 21 (10.4%) 32 (9.3%)
Other 2 (6.5%) 4 (10.3%)
Overall 113 (18.0%) 100 (16.3%)
Gestational age at birth (wk)
White, not Hispanic 25.9 (1.7) 25.9 (1.7)
Hispanic 25.8 (1.7) 25.7 (1.9)
African American 26.2 (2.0) 25.6 (2.1)
Other 25.6 (1.4) 26.1 (1.4)
Overall 26.0 (1.8) 25.9 (1.9)
Birthweight (g)
White, not Hispanic 785 (210) 784 (196)
Hispanic 772 (215) 802 (226)
African American 822 (277) 780 (265)
Other 729 (167) 813 (153)
Overall 792 (234) 787 (226) Figure 1. Study selection flow chart. *MAPPiNO study: (Askie
LM, Ballard RA, Cutter G, et al. Inhaled nitric oxide in preterm
infants: a systematic review and individual patient data meta-
analysis. BMC Pediatrics 2010;10(1):15.)

Table V. Continuous outcomes by trial

Overall mean (SD) Adjusted mean difference
Outcomes iNO Placebo Diff LCL UCL P value Interaction P value
PMA when ETT ceased (days)
Schreiber 219.49 (39.86) 215.46 (22.71) 4.09 −4.73 12.91 .364
NOCLD 237.66 (49.60) 240.49 (36.26) −2.83 −9.87 4.20 .430
NEWNO 225.41 (39.30) 223.33 (29.61) 2.26 −4.64 9.17 .520
Overall 230.34 (45.24) 230.30 (33.61) 0.02 −4.41 4.45 .993 .435
PMA when respiratory support ceased (days)
Schreiber 219.41 (40.55) 215.49 (22.83) 3.98 −5.11 13.07 .391
NOCLD 253.36 (50.03) 256.05 (36.53) −2.72 −9.86 4.42 .456
NEWNO 265.72 (58.94) 265.21 (41.07) 0.43 −8.79 9.65 .928
Overall 252.40 (54.36) 252.63 (40.19) −0.44 −5.42 4.55 .864 .529
PMA when discharged (days)
Schreiber 254.64 (39.33) 251.34 (40.07) 4.23 −6.91 15.37 .457
NOCLD 286.14 (52.17) 290.06 (41.47) −3.89 −11.5 3.72 .316
NEWNO 296.97 (52.28) 293.39 (34.40) 3.83 −5.05 12.71 .398
Overall 284.35 (52.18) 284.60 (41.73) −0.13 −5.24 4.97 .959 .366
Duration of oxygen (days)
Schreiber 29.75 (42.40) 29.99 (26.39) −0.14 −10.5 10.19 .978
NOCLD 58.47 (49.99) 61.78 (36.53) −3.34 −10.5 3.78 .357
NEWNO 86.34 (59.98) 86.02 (41.98) 0.29 −9.22 9.79 .953
Overall 64.60 (56.31) 66.40 (41.71) −1.52 −6.68 3.64 .564 .800
Duration of hospital stay (days)
Schreiber 68.60 (34.97) 74.51 (36.54) −5.97 −16.5 4.61 .269
NOCLD 106.94 (53.02) 111.35 (42.04) −4.44 −12.2 3.33 .263
NEWNO 116.73 (54.10) 113.93 (36.53) 3.35 −5.98 12.68 .481
Overall 104.53 (53.41) 107.05 (41.52) −2.13 −7.47 3.20 .433 .353

ETT, endotrachael tube; PMA, postmenstrual age.

39.e1 Askie et al
February 2018 ORIGINAL ARTICLES

Table VI. Continuous outcomes by race/ethnicity

Outcomes Overall mean (SD) Adjusted mean difference
Race iNO Placebo Diff LCL UCL P value Interaction P value
PMA when ETT ceased (days)
White, not Hispanic 231.71 (48.47) 235.87 (33.55) −3.99 −11.1 3.09 .269
Hispanic 226.58 (31.35) 227.62 (29.94) −0.48 −9.67 8.70 .918
African American 228.90 (46.16) 227.21 (35.34) 1.69 −6.16 9.55 .673
Other 236.62 (36.50) 220.50 (27.92) 12.08 −2.02 26.18 .093
Overall 230.34 (45.24) 230.30 (33.61) −0.18 −4.58 4.23 .937 .123
PMA when off respirator (days)
White, not Hispanic 257.12 (57.67) 261.42 (38.76) −4.95 −13.0 3.06 .226
Hispanic 252.22 (43.62) 250.53 (33.12) 2.12 −8.08 12.32 .684
African American 245.24 (54.72) 243.38 (44.12) 2.57 −6.09 11.22 .561
Other 252.39 (40.16) 257.28 (29.69) −7.56 −24.5 9.34 .381
Overall 252.40 (54.36) 252.63 (40.19) −0.89 −5.83 4.06 .725 .326
PMA when discharged (days)
White, not Hispanic 289.92 (56.16) 292.15 (38.11) −2.26 −10.4 5.87 .586
Hispanic 287.05 (45.01) 280.80 (38.55) 6.30 −5.51 18.12 .296
African American 274.52 (50.07) 278.10 (47.54) −2.97 −11.6 5.70 .502
Other 287.30 (31.07) 284.31 (25.28) 0.36 −14.4 15.11 .962
Overall 284.35 (52.18) 284.60 (41.73) −0.54 −5.59 4.51 .834 .980
Duration of oxygen (days)
White, not Hispanic 67.32 (59.00) 71.43 (41.45) −4.79 −12.9 3.35 .249
Hispanic 67.60 (46.21) 66.42 (34.56) 0.90 −9.12 10.92 .860
African American 58.75 (58.32) 59.61 (45.82) 0.37 −8.78 9.52 .937
Other 65.84 (37.56) 72.61 (31.32) −8.18 −24.1 7.72 .313
Overall 64.60 (56.31) 66.40 (41.71) −1.71 −6.84 3.42 .514 .391
Duration of hospital stay (days)
White, not Hispanic 108.83 (57.41) 111.92 (39.26) −2.93 −11.4 5.58 .499
Hispanic 109.99 (44.47) 103.94 (39.52) 6.37 −5.58 18.33 .296
African American 94.79 (51.72) 103.25 (46.65) −7.75 −16.6 1.12 .087
Other 108.67 (34.94) 104.76 (24.22) −1.73 −16.4 12.91 .817
Overall 104.53 (53.41) 107.05 (41.52) −2.33 −7.62 2.97 .389 .705

ETT, endotrachael tube; PMA, postmenstrual age.

Race Effects of Inhaled Nitric Oxide in Preterm Infants: An Individual Participant Data Meta-Analysis 39.e2