Childhood Blood Pressures

Pharmacological Treatment of Arterial Hypertension

in Children and Adolescents
A Network Meta-Analysis
Jacopo Burrello,* Elvira M. Erhardt,* Gaelle Saint-Hilary, Franco Veglio, Franco Rabbia,
Paolo Mulatero, Silvia Monticone,† Fabrizio D’Ascenzo†

Abstract—Pharmacological treatment is indicated in children and adolescents with hypertension unresponsive to lifestyle
modifications, but there is not enough evidence to recommend 1 class of antihypertensive drugs over others. We performed
a network meta-analysis to compare the results of available randomized clinical trials on pharmacological treatment of
pediatric hypertension. From a total of 554 potentially relevant studies, 13 randomized placebo-controlled clinical trials
enrolling ≥50 patients and a follow-up ≥4 weeks were included. The reduction of systolic blood pressure (SBP) and
diastolic BP (DBP) after treatment were the coprimary end points. A total of 2378 pediatric patients, with a median age
of 12 years, were included in the analysis. After a median follow-up of 35 days, lisinopril and enalapril were found to
be superior to placebo in reducing SBP and DBP, whereas only for DBP, losartan was found to be superior to placebo
and lisinopril and enalapril were found to be superior to eplerenone. Angiotensin-converting enzyme inhibitors and
angiotensin receptor blockers were associated with a greater SBP and DBP reduction compared with placebo, likewise
the mineralocorticoid receptor antagonist was inferior to angiotensin-converting enzyme inhibitors in DBP reduction. The
analysis was adjusted for study-level mean age, percentage of women, mean baseline blood pressure, and mean weight,
only the latter significantly affected DBP reduction. According to the present analysis, angiotensin-converting enzyme
inhibitors and angiotensin receptor blockers could represent the best choice as antihypertensive treatment for pediatric
hypertension. However, because of the paucity of available data for the other classes of antihypertensive drugs, definitive
conclusions are not allowed and further randomized controlled trials are warranted.  (Hypertension. 2018;72:306-313.
DOI: 10.1161/HYPERTENSIONAHA.118.10862.) Online Data Supplement •
Downloaded from http://ahajournals.org by on July 31, 2018

Key Words: adolescent ◼ angiotensin-converting enzyme inhibitors ◼ blood pressure

◼ hypertension ◼ network meta-analysis

I n children and adolescents, given the lack of outcome data,

the definition of hypertension is based on the normal distri-
bution of blood pressure (BP) in healthy children. For patients
aged up to 131 or up to 16 years,2 hypertension is defined as
systolic BP (SBP) and/or diastolic BP (DBP) ≥95th percentile
for healthy children of the same sex, age, and height on at
least 3 separate measurements. Recent evidence indicates that
childhood hypertension is not uncommon, with a prevalence,
in Europe, comprised between 2.2% and 13%2 that increases
up to 22% in overweight or obese children.3 Moreover, with
the revised definition of BP categories, in agreement with the
American Heart Association/American College of Cardiology
guideline4 and the introduction of the new normative BP
tables, based exclusively on BP recordings from normal
weight children,1 a significant number of subjects are expected
to be reclassified as affected by arterial hypertension.
The relationship between hypertension and cardiovascular
risk is well established in adults,5 and BP lowering has been
demonstrated to reduce the excess risk in long-term random-
ized controlled trials (RCTs).6 Similar evidence is lacking for
children, but childhood hypertension is significantly associ-
ated with target organ damage7,8 and antihypertensive drugs,
effective in reducing BP, lead to the regression of left ventricu-
lar hypertrophy and reduction of intima-media thickness and
microalbuminuria.9–12 Furthermore, it has been reported a sig-
nificant association between BP levels in late adolescence and
cardiovascular mortality several decades later in a Swedish
cohort of >1 million male conscripts13 and in a cohort of
18 881 male university students.14 Similarly, childhood hyper-
tension was significantly associated with the rate of premature
death (before 55 years of age) from endogenous causes in a
cohort of 4857 Native Americans.15

Received January 18, 2018; first decision February 2, 2018; revision accepted June 6, 2018.
From the Division of Internal Medicine and Hypertension (J.B., F.V., F.R., P.M., S.M.) and Division of Cardiology (F.D.A.), Department of Medical
Sciences, University of Turin, Italy; and Department of Mathematical Sciences G. L. Lagrange, Polytechnic University of Turin, Italy (E.M.E., G.S.H.).
*These authors contributed equally to this work.
†These authors contributed equally to this work.
The online-only Data Supplement is available with this article at https://www.ahajournals.org/journal/hyp/doi/suppl/10.1161/HYPERTENSIONAHA.
118.10862.
Correspondence to Silvia Monticone, Division of Internal Medicine and Hypertension, Department of Medical Sciences, University of Turin, Via Genova
3, 10126 Torino, Italy. E-mail silvia.monticone@unito.it
© 2018 American Heart Association, Inc.
Hypertension is available at https://www.ahajournals.org/journal/hyp DOI: 10.1161/HYPERTENSIONAHA.118.10862

306
 Burrello et al   Hypertension Management in Children   307
According to the European and American Society guide-
lines, pharmacological treatment is indicated in hypertensive
children unresponsive to lifestyle modifications, as well as in
all children with symptomatic hypertension, secondary hyper-
tension, target organ damage, chronic kidney disease, or dia-
betes mellitus.1,2 Pharmacological therapy should be started
with the lowest dose of a single drug then up-titrated until BP
control or until the maximum recommended dose is reached.
When target BP is not achieved with a single agent, a second
drug should be added and up-titrated.1,2 However, the lack of
head-to-head pharmacological trials does not allow to provide
strong recommendations for any class of antihypertensive
medications over the others, and the choice of a particular drug
is usually driven by hypertension underlying pathogenesis (in
case of secondary hypertension), concurrent disorders, side
effects, and clinician’s preference.1,2 Network meta-analysis
(NMA), allowing an integrated analysis of RCTs that com-
pare different drugs head-to-head or with placebo whereas
fully respecting randomization,16 can significantly influence
and inform treatment recommendations4,17,18 whereas network
meta-regression (NMR) explores the impact of moderators/
effect modifiers in the evidence network.19,20
Therefore, the aim of our study is to compare the effective-
ness in SBP and DBP reduction of different antihypertensive
medications in children, combining the data from all the avail-
able randomized placebo-controlled clinical trials, through a
NMA and NMR approach.
Methods
Downloaded from http://ahajournals.org by on July 31, 2018
All supporting data are available within the article and its online sup-
plementary files. The study protocol can be found online at https://
www.crd.york.ac.uk/prospero/display_record.php?RecordID=84462.
Search Strategies, Eligibility Criteria, and
Information Sources
We searched MEDLINE and Cochrane Library for RCTs on treat-
ment of pediatric hypertension. We restricted the search to human
studies, clinical studies, controlled trials, or randomized trials. We
used the keywords: child/children/child*, pediatric, hypertension/
hypertensives/hypertens*, treatment/treat*, and therapy, as well as
additional text words in combination with an established search strat-
egy for MEDLINE/PubMed. We also hand searched in references of
identified studies and recent guidelines and reviews on the topic.
Study Selection and Data Collection Process
Study selection was performed by 2 independent reviewers (J. Burrello
and F. Rabbia), with divergences solved by discussion. Retrieved
citations were first screened for relevance at the title/abstract level,
then shortlisted studies were examined in full text. Studies were con-
sidered suitable for inclusion if (1) randomized trials; (2) placebo-
controlled; (3) >50 patients were enrolled; and (4) follow-up was ≥4
weeks. We decided to choose this cutoff for the follow-up because
for angiotensin-converting enzyme (ACE) inhibitors and angiotensin
receptor blocker (ARBs), the maximum antihypertensive efficacy is
achieved after 2 to 4 weeks of treatment.12,21 Studies were excluded if
(1) inclusion criteria were not all fulfilled; (2) missing values or data;
(3) duplicated reports; and (4) studies were conducted in a selected
cohort of patients with specific clinical characteristics (eg, renal dis-
ease with proteinuria, aortic coarctation).
Clinical Data and End Points
Data abstraction and study appraisal were performed by 2 indepen-
dent reviewers (J. Burrello and E.M. Erhardt). Extracted data com-
prised study-level summary measures of age, sex, weight, body mass
index, duration of follow-up, baseline SBP and DBP, and follow-up
SBP and DBP for each selected study. For 1 study only δ-DBP (ie,
change from baseline, follow-up DBP−baseline DBP) was available.
δ-SBP (ie, change from baseline, follow-up SBP−baseline SBP) and
δ-DBP were evaluated as coprimary end points.
Arms
The following drugs and pharmacological classes were considered:
(1) ACE inhibitors (enalapril, fosinopril, and lisinopril); (2) ARBs
(losartan, olmesartan, telmisartan, and valsartan); (3) calcium chan-
nel blockers (amlodipine and felodipine); (4) β-blockers (metopro-
lol; bisoprolol+hydrochlorothiazide, which was included only in
main NMA); and (5) mineralocorticoid receptor antagonist (MRA:
eplerenone).
Data Analysis
Continuous variables are reported as mean or median (25th–75th
percentiles, interquartile range). Categorical variables are expressed
as n/N (%). The continuous end points were SBP and DBP and the
differences between the treatment groups were measured in terms of
difference between mean changes from baseline. All the 13 (12 for
SBP) included studies compared 1 treatment versus placebo (Table;
Figure 1A). Four studies (3 for SBP) compared different dosages of
the same drug to 1 placebo arm and their differences were pooled
into 1 arm each, weighting by the inverse of the variance of the dif-
ferences. In 2 studies the included patients were subdivided into 2
cohorts (by race and by obesity, respectively). Further, 4 studies com-
pared different dosages of the same drug, each with specific placebo
arm; hence, for analysis purpose, they were considered as separate
studies.
First, for each treatment and for each class of drug—whether they
had been tested in >1 study—the direct estimates were obtained using
standard pairwise meta-analysis. Standard pairwise meta-analyses
were conducted in a frequentist framework applying random effect
models. The between-study heterogeneity was assessed via visual
inspection of the forest plots with their 95% confidence intervals,
the I2 statistic, and the P value of the Cochrane Q test (Table S1 in
the online-only Data Supplement); the between-study variance was
assessed using the τ2 statistic (Table S2). Second, NMAs were per-
formed in a Bayesian framework, using a random effect model with
the assumption of homogeneous between-study variance.19,35 Finally,
NMR were conducted to assess the impact of different covariates on
each comparison versus placebo, assuming the covariates do not have
any impact on pairwise comparisons between treatments.36
All parameters were given minimally informative priors. The
posterior distributions of the parameters were obtained using Markov
Chain Monte Carlo methods, running 2 chains. The chains were
thinned to reduce autocorrelations. For each chain, the first 100 000
(for the NMR, 120 000) Monte Carlo iterations were run to reach
convergence and disregarded, then 120 000 (for the NMR, 200 000)
further simulations per chain were run to estimate the parameters.
The autocorrelation of the chains was checked through inspection
of autocorrelation plots. A visual inspection of the history plots was
used to assess the convergence of the chains. The between-study vari-
ance τ2 was checked (Table S2).
For each analysis of each end point, the table of differences
between the treatments and the rank probability diagram were ex-
amined. In Figures 2 and 3 and Figure S1, S2, and S13B in the online-
only Data Supplement, the median and the credibility interval of all
possible comparisons between treatments are presented. The rank
probability diagram is a histogram providing a visual interpretation of
the rankings of the treatments. For each treatment, the probability to
be the best, the second best, etc. is depicted (Figures S3 through S12).
For analysis purpose, the studies were further divided in low age
range studies and high age range depending on whether the mean age
of the included patients was ≤12 or >12 years, as detailed in Table.
An extended description of the statistical methods is reported in the
online-only Data Supplement.
The publication bias was assessed using funnel plot and Egger
test.
 308  Hypertension   August 2018

Table.  Included Studies and Patients Characteristics

Total Placebo/ Baseline

Selected Study no. of Treatment Follow- Treatment Age Baseline SBP, DBP, Sex (%
Study Drug Design Patients (patients) Up (d) Strategy Range BMI, kg/m2 Weight, kg Age, y mm Hg mm Hg Women)

Meyers, et Valsartan Placebo 245 128/117 28 Dose/kg Low 27.0 62 11.3 132 78 39
al22 withdrawal

Wells, et al23 Valsartan Placebo 245 122/123 28 Dose/kg Low NA 66 11.4 133 78 40
withdrawal

Li, et al24 Eplerenone Placebo 304 137/167 70 Standard Low/ 28.5 71 12.3 127 71 37
concurrent dose high

Hazan, et Olmesartan Placebo 286 141/145 35 Dose/kg High 28.1 71 12.3 130 78 41
al25 withdrawal

Wells, et al26 Telmisartan Placebo 73 14/59 28 Dose/kg High 28.8 75 14.0 132 79 43
concurrent

Batisky, et Metoprolol Placebo 140 23/117 28 Dose/kg High 28.0 NA NA 132 78 30

al27 concurrent

Shahinfar, Losartan Placebo 164 87/77 36 Dose/kg Low/ NA 59 12 130 89 44

et al28 withdrawal high

Flynn, et al29 Amlodipine Placebo 268 90/178 56 Standard High 26.6 66 12.1 138 74 NA
concurrent dose

Li, et al30 Fosinopril Placebo 221 115/106 42 Dose/kg High 29.7 74 12.1 134 72 34
withdrawal

Soffer, et al31 Lisinopril Placebo 104 51/53 28 Dose/kg High NA 56 NA 129 90 35

withdrawal

Trachtman, Felodipine Placebo 133 35/98 28 Standard High NA 77 12.1 NA NA 40

et al32 concurrent dose

Sorof, et al33 Bisoprolol+ Placebo 94 32/62 98 Standard High 28.3 77 13.9 134 83 43
HCTZ concurrent dose

Wells, et al34 Enalapril Placebo 101 50/51 28 Dose/kg Low NA 53 11.6 129 90 42
Downloaded from http://ahajournals.org by on July 31, 2018

withdrawal

Network NA NA 2378 1025/1353 35 NA NA 28.0 66.4 12.1 130 83 40

meta- (28–56) (27.0–28.5) (56.0–73.4) (11.8–12.3) (128.0–133.7) (74.2–88.1) (34.4–44.0)
analysis*
RCTs included in the network meta-analysis. Baseline weight, age, baseline SBP, and DBP are reported as weighted means for the selected studies and as median
and interquartile range (25th–75th percentiles, IQR) for network meta-analysis data. Age range was defined low if age was ≤12 y or high if age was >12 y. BMI indicates
body mass index; HCTZ, hydrochlorothiazide; NA, not applicable/not available; and RCTs, randomized controlled trials.
*Reports patients characteristics of our analysis.

The present research was performed following the Preferred
Reporting Items for Systematic reviews and Meta-Analyses amend-
ment37 to the Quality of Reporting of Meta-analyses statement,38 The
Cochrane Collaboration, and Meta-analysis Of Observational Studies
in Epidemiology.39
Internal Validity and Quality Appraisal
Quality of included trials was explored according to Cochrane state-
ments19,36 selection bias (allocation and random sequence generation),
performance bias (blinding of participants and personnel), detection
bias (blinding of outcome assessment), and attrition bias (incomplete
outcome data).
Results
The literature search resulted in the identification of 554
potentially relevant studies: 494 were excluded because they
did not fulfill all the inclusion criteria; of the 60 full text
reports assessed, 47 were excluded because of missing data
(n=9), duplicated reports (n=3), selected cohorts (n=8), not
randomized (n=16), ongoing studies (n=1), or retrospective
design (n=10). Thirteen studies (all RCTs comparing an
antihypertensive drug versus placebo), including a total of
2378 patients (1025 in the placebo group and 1353 in active
treatment) with a median follow-up of 35 days (28–56) were
eligible for the final analysis (Figure 1A).22–34 As detailed
in Table S3, in 4 studies patients affected by secondary
forms of hypertension were excluded, in 6 studies they were
excluded, but not systematically (detailed description of the
exclusion criteria is provided), whereas in 3 studies the pres-
ence of secondary hypertension was not investigated. The
main clinical and biochemical characteristics of the included
patients are summarized in Table, 40% (34.4–44) were
women, the median age of the resulting cohort was 12.1
(11.8–12.3) years, the median weight was 66.4 kg (56–73.4),
the median body mass index was 28.0 kg/m2 (27.0–28.5) and
the median baseline SBP and DBP were, respectively, 130
and 83 mm Hg (128–134; 74–88).
For 7 studies, the design was randomized double-blind
placebo withdrawal (all patients received open-label treat-
ment, then they were randomized to continue the active treat-
ment or to receive a placebo) and randomized double-blind
placebo concurrent for 6 studies (after randomization, 1 arm
was given the treatment, the other a placebo). The treatment
 Burrello et al   Hypertension Management in Children   309

Figure 1.  Study design. A, Flowchart of study selection. B, Evidence network diagram. The size of each treatment node is proportional to the number of
treated patients (shown in round brackets). The randomized controlled trials comparing the drug with the placebo is reported on the arrow. Gray boxes group
drugs according to their class: β-Bs indicates β-blockers; ACE inhibitors, angiotensin-converting enzyme inhibitors; ARBs, angiotensin receptor blockers;
CCBs, calcium channel blockers; HCTZ, hydrochlorothiazide; and MRA, mineralocorticoid receptor antagonist.

strategy was different: in 4 studies (including a total of 799
patients), a weight-independent standard dose was adminis-
tered, in the other 9 studies (for a total of 1579 patients) a dose
per kg strategy. For all the selected studies, the evaluated end
points were SBP and DBP at baseline and after treatment (for
placebo-concurrent studies) or after treatment withdrawal (for
placebo-withdrawal studies).
Downloaded from http://ahajournals.org by on July 31, 2018
both SBP (7.7 mm Hg [1.5–13.2] and 7.9 mm Hg [2–14],
respectively) and DBP reduction (6.6 mm Hg [1.7–11.0]
and 6.3 [1.6–11.0], respectively), whereas no significant
differences were found in the NMA of studies in which
a weight-independent standard dose strategy was adopted
(Figure S1B).
Among studies, including patients with a mean age

The evidence network for the main analyses (Figure 1B)
comprised 13 RCTs and 12 different drugs: 3 ACE inhibitors
(enalapril,34 fosinopril,30 and lisinopril31), ARBs (losartan,28
olmesartan,25 telmisartan,26 and valsartan22,23), 2 calcium chan-
nel blockers (amlodipine29 and felodipine32), 1 MRA (eplere-
none24), 1 β-blocker (metoprolol27), and an association of a
β-blocker with a thiazide diuretic (bisoprolol+hydrochloroth
iazide33).
In the NMA for single drug (Figure 2), lisinopril and
enalapril were associated with a significant reduction in SBP
and DBP compared with placebo (7.8 mm Hg [2.3–12.8]/6.6
mm Hg [2.5–10.5] and 7.8 mm Hg [2.5–13.5]/6.2 mm Hg [2–
10.6], respectively). Lisinopril and enalapril were superior
also to eplerenone in DBP reduction (6.5 mm Hg [1.1–11.6]
and 6.1 mm Hg [0.7–11.8], respectively). Finally, a signifi-
cant DBP reduction was observed in losartan-treated patients
compared with placebo (3.7 mm Hg [0.3–7.5]). Consistently,
in NMA by pharmacological classes of drugs (Figure 3),
ACE inhibitors and ARBs were associated with a signifi-
cant reduction in SBP and DBP (6.5 mm Hg [3.7–9.7]/4.4
mm Hg [2.2–7.3] and 3.3 mm Hg [1.2–5.8]/3.1 mm Hg [1.4–
5.0], respectively) compared with placebo and ACE inhibi-
tors were superior to MRA in DBP reduction (4.3 mm Hg
[0.2–9.1]).
To assess the impact of treatment strategy and age on
BP reduction, we performed 2 subgroup NMAs. In agree-
ment with main analysis, in the NMA, including only the
studies with a dose per kg strategy (Figure S1A), lisino-
pril and enalapril confirmed their superiority to placebo in
>12 years, only lisinopril was associated with a significant
reduction in SBP and DBP compared with placebo (7.9
[1.1–13.5]/6.55 [2.2–10.7]; Figure S2A). In NMA of studies,
including patients with a mean age ≤12 years, (Figure S2B),
enalapril was associated with a reduction in SBP and DBP
compared with placebo (7.7 mm Hg [3–12.4]/6.3 mm Hg [1–
11.4]); for SBP only, losartan was superior to placebo (9.3
mm Hg [2.5–15.8]) and enalapril was superior to valsartan
(5.8 mm Hg [0.1–11.3]).
NMR analyses with covariate adjustments based on study-
level summary measures of age, sex, baseline weight, and
baseline BP were performed to correct for potentially con-
founding variables: a significant effect was found for mean
weight on DBP, whereas the other evaluated parameters did
not significantly affect the end points. The NMR on DBP
reduction adjusted by study mean weight (64.3±10.8 kg) dem-
onstrated that an increase in body weight of 1 kg is associated
with an increased difference versus placebo of 0.181 (0.043–
0.327) mm Hg (Table S4). The regression lines showed that
the higher was the mean weight in the study the better was the
response to antihypertensive treatment (Figure S13A). After
correcting for study mean weight, lisinopril and enalapril were
associated with an even greater reduction in DBP (7.9 mm Hg
[4.1–11.7] and 8.5 mm Hg [4.3–12.7], respectively) compared
with placebo, considering that these 2 cohorts of patients
displayed the lowest mean weight at baseline (56 and 53 kg,
respectively).
The mean age, baseline weight, and baseline BP and
women percentage of studies included in the NMR analysis
 310  Hypertension   August 2018
Downloaded from http://ahajournals.org by on July 31, 2018
Figure 2.  Systolic blood pressure (SBP) and diastolic blood pressure
(DBP) reduction from main network meta-analysis. The figure (Continued )
for SBP and DBP are reported in Table S4, together with their
effect and 95% credibility interval. We performed also rank
probability histogram analyses, which supported our results
(Figures S3 through S12) and standard pairwise meta-anal-
yses for each pairwise comparison to obtain direct estimates
of the treatment difference. Forest plots of the direct compari-
sons can be found in Figures S14 through S23.
All the included studies were randomized, with an overall
low risk of bias (Table S5; Figures S24 and S25). The Eggers
test for publication bias was not significant in any of the cases
(P value, 0.24 for SBP and P value, 0.10 for DBP).
Discussion
The prevalence of arterial hypertension in children and ado-
lescents, together with overweight and obesity, has increased
markedly in recent years: recent estimates indicate that up to
13% of children in Europe display elevated BP.2 Strong evi-
dence indicates that childhood hypertension is associated with
cardiac organ damage7,8 and represents a strong risk factor for
the development of hypertension in adulthood.40–42 Moreover,
the paucity of clinical data supporting the efficacy of anti-
hypertensive medications and the lack of strong evidence
to recommend 1 class of antihypertensive medications over
the others, hampers the appropriate management of pediatric
hypertension.
The main finding of our NMA is that, in children and ado-
lescents, ACE inhibitors and ARBs are superior to placebo in
the reduction of both SBP and DBP. Additionally, overcom-
ing the lack of head-to-head trials through a NMA approach,
we demonstrated that ACE inhibitors are superior to MRA in
DBP reduction, whereas no significant differences were found
among the other antihypertensive medications.
The results of the present analysis support the common use
of ACE inhibitors in clinical practice for the management of
pediatric hypertension43; moreover, a recent systematic review
of the literature showed that lisinopril and enalapril produced
the greatest absolute BP reduction compared with other anti-
hypertensive drugs.44 ACE inhibitors approved by the Food
and Drug Administration (FDA) are benazepril, enalapril,
fosinopril, and lisinopril, but limited to children ≥6 years (≥1
month for enalapril) of age with a glomerular filtration rate
≥30 mL/min.45 For this indication, FDA approved enalapril,
fosinopril, and lisinopril, respectively, in 2002, 2004, and
2003 after RCTs demonstrating their safety and efficacy in
children30,31,34; benazepril was approved by FDA even if no
RCTs were available in the literature.45
In our NMA we demonstrated that ARBs, pooled together,
were superior to placebo, whereas looking at the single drug
Figure 2 Continued. shows the average reduction (mm Hg) after treatment
(and its 95% credibility interval) of SBP (δ-SBP, above the diagonal
identified by drug names) and DBP (δ-DBP, below the diagonal identified
by the names of drugs). For SBP (above the diagonal) a positive value
identifies a reduction of blood pressure (BP) in favor of the column-defining
treatment; a negative value identifies a reduction of BP in favor of the row-
defining treatment. For DBP (below the diagonal) a positive value identifies
a reduction of BP in favor of the row-defining treatment; a negative value
identifies a reduction of blood pressure (BP) in favor of the column-
defining treatment. In bold the statistically significant values (differences
are considered as statistically significant when the 0 is not included in the
95% credibility interval). HCTZ indicates hydrochlorothiazide; and NA, not
available.
 Burrello et al   Hypertension Management in Children   311

Figure 3.  Network meta-analysis by classes of drugs. The figure shows the average reduction (mm Hg) after treatment (and its 95% credibility) of systolic
blood pressure (SBP; δ-SBP, above the diagonal identified by drug names) and diastolic blood pressure (DBP; δ-DBP, below the diagonal identified by the
names of drugs). For SBP (above the diagonal) a positive value identifies a reduction of blood pressure (BP) in favor of the column-defining treatment; a
negative value identifies a reduction of BP in favor of the row-defining treatment. For DBP (below the diagonal) a positive value identifies a reduction of BP in
favor of the row-defining treatment; a negative value identifies a reduction of BP in favor of the column-defining treatment. In bold the statistically significant
values (differences are considered as statistically significant when the 0 is not included in the 95% credibility interval). Angiotensin receptor blockers (ARBs;
losartan, olmesartan, telmisartan, valsartan); β-blockers (β-Bs; metoprolol); angiotensin-converting enzyme (ACE) inhibitors (enalapril, fosinopril, and lisinopril);
calcium channel blockers (CCBs; amlodipine and felodipine); mineralocorticoid receptor antagonists (MRA; eplerenone); bisoprolol+hydrochlorothiazide was
excluded from network meta-analysis by classes of drugs.

analysis, only losartan displayed a significant DBP reduction
versus placebo, but no difference was found for SBP. This
result could be explained by the increased precision of the
pooled analysis and the favorable trend observed for the other
ARBs. ARBs approved by the FDA are candesartan, losartan,
olmesartan, and valsartan.45
Among MRAs, eplerenone was studied on a large cohort
of 304 hypertensive children, but a significant BP reduction
Downloaded from http://ahajournals.org by on July 31, 2018
significant effect was found only for study mean weight in DBP
reduction. Even if in literature obesity is associated with a worse
response to antihypertensive treatment,22 in our NMR analysis
the increase in study mean weight was associated with a greater
BP drop. It should be noted that because individual patient data
were not available, our NMR was based on summary measures at
the study level and the association between weight and BP reduc-
tion detected at the study level in our analysis may not reflect the

was demonstrated only for the highest dose and only for SBP24
and the FDA did not approve it. In our analysis, eplerenone
was the only drug that resulted inferior to another in head-
to-head comparison (lisinopril and enalapril), whereas there
was no significant difference versus placebo. Similarly, MRAs
resulted inferior to ACE inhibitors in DBP reduction in our
analysis by classes of drugs.
As reported in literature, the absolute BP-lowering effect may
be influenced by several factors, including age, sex, weight, and
severity of baseline hypertension. Age significantly affects treat-
ment adherence: compliance in adolescence is significantly lower
than in prepubertal age and associated with a poor outcome46;
obesity is associated with a worse response to antihypertensive
therapy,22 whereas a higher baseline BP predicts a greater BP
reduction after treatment.44 Finally, female patients seem to have
a better adherence to pharmacological treatment (as demonstrated
for other medical conditions).47 In our study, we investigated the
effect of these factors through subgroup NMAs (by treatment
strategy and age) and NMR. Subgroup NMAs confirmed the
superiority of lisinopril and enalapril (administered using a dose
per kg strategy) in both SBP and DBP reduction compared with
placebo, the former in a population with a mean age >12 years
and the latter in a population with a mean age ≤12 years. Despite
the importance of considering the weight of patients when anti-
hypertensive treatment is initiated,1,2 the results of the present
analysis do not allow to determine whether the dose regimen has
an impact on treatment effect because treatments with fixed dose
and treatments with dose per kg are not the same.
In our NMR analysis, age, sex, and baseline BP did not
significantly affect SBP or DBP reduction at the study level; a
individual level effect modification of that covariate.19,48–51
This is the first study comparing head-to-head the efficacy
in BP reduction of different antihypertensive medications on
a relatively high number of pediatric patients, demonstrating
that only 2 classes of drugs, ACE inhibitors and ARBs, per-
form better than placebo in this special subpopulation.
Our study has potential limitations. First, the paucity of
clinical trials investigating other classes of antihypertensive
medications, other than the newer ACE inhibitors and ARBs,
could have influenced the results; second, heterogeneity, small
size, and short follow-up of RCTs included in the analysis, as
well as the lack of studies comparing antihypertensive drugs of
different classes between them, should be taken into account.
Third, patients affected by secondary forms of hypertension,
which benefit from specific and targeted therapies, were not
systematically excluded in all studies. However, because
patients with stage 3, or higher, chronic kidney disease52 (the
most common secondary cause of hypertension in children1),
were not included in 10 out of 13 studies, it is extremely
unlikely that this heterogeneity could have affected our final
results. Moreover, our primary end point was BP reduction
from baseline, but there are no data on target organ damage
prevention/regression or other strong outcome predictors.
Further RCTs are warranted to definitively confirm
whether ACE inhibitors and ARBs could actually represent
the best choice for the management of pediatric hypertension
independently from the main confounding factors, including
age, sex, weight, and baseline BP.
Notwithstanding the reported limitations, the findings
from the present meta-analysis represent a comprehensive
 312  Hypertension   August 2018
evidence base to guide the choice about pharmacological treat-
ment for arterial hypertension in children and adolescents.
Perspectives
Arterial hypertension affects up to 13% of the pediatric popu-
lation in Europe,2 target organ damage associated with elevated
BP starts in childhood,7,8 and several studies demonstrated the
evidence of BP tracking from childhood into adulthood.40 A
well-timed diagnosis of childhood hypertension, including the
diagnostic workup for secondary forms, is of outmost impor-
tance to start the appropriate treatment and to prevent the
development of cardiovascular morbidity associated with high
BP levels. Clinicians have a wide range of pharmacological
options for the management of pediatric hypertension, how-
ever, strong long-term outcome data are lacking for children
and further head-to-head RCTs are required to fully elucidate
the comparative benefits and safety profile of each single drug
and combination therapies and to demonstrate the effects on
prevention and regression of target organ damage.
Acknowledgments
All the authors contributed extensively to the work presented in this
article. S. Monticone, P. Mulatero, and F. D’Ascenzo conceived the
study. J. Burrello, E.M. Erhardt, and G. Saint-Hilary performed the
literature search, analyzed the data, and wrote the article. F. Veglio,
F. Rabbia, and G. Saint-Hilary contributed to study protocol and
key data interpretation. P. Mulatero, F. Rabbia, F. Veglio, and S.
Monticone critically revised the article. The corresponding author
had full access to all of the data and the final responsibility to submit
for publication.
Downloaded from http://ahajournals.org by on July 31, 2018
Sources of Funding
E.M. Erhardt received funding from the European Union’s
Horizon 2020 research and innovation programme under the Marie
Sklodowska-Curie grant agreement no. 633567.
Disclosures
G. Saint-Hilary research is funded by Servier and she is a consultant
for AstraZeneca. The other authors report no conflicts.
References
1. Flynn JT, Kaelber DC, Baker-Smith CM, et al. Clinical practice guideline
for screening and management of high blood pressure in children and ado-
lescents. Pediatrics. 2017;140:e20171904.
2. Lurbe E, Agabiti-Rosei E, Cruickshank JK, et al. 2016 European Society
of Hypertension guidelines for the management of high blood pressure in
children and adolescents. J Hypertens. 2016;34:1887−1920.
3. Flechtner-Mors M, Neuhauser H, Reinehr T, Roost HP, Wiegand S,
Siegfried W, Zwiauer K, Molz E, Holl RW; APV Initiative and the BMBF
Competence Network Obesity. Blood pressure in 57,915 pediatric patients
who are overweight or obese based on five reference systems. Am J
Cardiol. 2015;115:1587–1594. doi: 10.1016/j.amjcard.2015.02.063.
4. Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/
ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the pre-
vention, detection, evaluation, and management of high blood pressure in
adults: a report of the American College of Cardiology/American Heart
Association Task Force on Clinical Practice Guidelines. Hypertension.
2018;71:e13–e115. doi: 10.1161/HYP.0000000000000065.
5. Rapsomaniki E, Timmis A, George J, Pujades-Rodriguez M, Shah AD,
Denaxas S, White IR, Caulfield MJ, Deanfield JE, Smeeth L, Williams B,
Hingorani A, Hemingway H. Blood pressure and incidence of twelve car-
diovascular diseases: lifetime risks, healthy life-years lost, and age-spe-
cific associations in 1·25 million people. Lancet. 2014;383:1899–1911.
doi: 10.1016/S0140-6736(14)60685-1.
6. Mancia G, Fagard R, Narkiewicz K, et al. 2013 ESH/ESC guidelines for
the management of arterial hypertension: the task force for the manage-
ment of arterial hypertension of the European Society of Hypertension
(ESH) and of the European Society of Cardiology (ESC). Eur Heart J.
2013;34:2159–2219. doi: 10.1093/eurheartj/eht151.
7. Sorof JM, Cardwell G, Franco K, Portman RJ. Ambulatory blood pressure
and left ventricular mass index in hypertensive children. Hypertension.
2002;39:903–908.
8. Litwin M, Niemirska A, Sladowska J, Antoniewicz J, Daszkowska J,
Wierzbicka A, Wawer ZT, Grenda R. Left ventricular hypertrophy and
arterial wall thickening in children with essential hypertension. Pediatr
Nephrol. 2006;21:811−819.
9. Litwin M, Niemirska A, Sladowska-Kozlowska J, Wierzbicka A, Janas
R, Wawer ZT, Wisniewski A, Feber J. Regression of target organ damage
in children and adolescents with primary hypertension. Pediatr Nephrol.
2010;25:2489–2499. doi: 10.1007/s00467-010-1626-7.
10. Seeman T, Gilík J, Vondrák K, Simková E, Flögelová H, Hladíková M,
Janda J. Regression of left-ventricular hypertrophy in children and ado-
lescents with hypertension during ramipril monotherapy. Am J Hypertens.
2007;20:990–996. doi: 10.1016/j.amjhyper.2007.03.009.
11. Assadi F. Effect of microalbuminuria lowering on regression of left ven-
tricular hypertrophy in children and adolescents with essential hyperten-
sion. Pediatr Cardiol. 2007;28:27−33.
12. Kassler-Taub K, Littlejohn T, Elliott W, Ruddy T, Adler E. Comparative
efficacy of two angiotensin II receptor antagonists, irbesartan and
losartan in mild-to-moderate hypertension. Irbesartan/Losartan Study
Investigators. Am J Hypertens. 1998;11(4 pt 1):445–453.
13. Sundström J, Neovius M, Tynelius P, Rasmussen F. Association of blood
pressure in late adolescence with subsequent mortality: cohort study of
Swedish male conscripts. BMJ. 2011;342:d643.
14. Gray L, Lee IM, Sesso HD, Batty GD. Blood pressure in early adult-
hood, hypertension in middle age, and future cardiovascular disease
mortality: HAHS (Harvard Alumni Health Study). J Am Coll Cardiol.
2011;58:2396–2403. doi: 10.1016/j.jacc.2011.07.045.
15. Franks PW, Hanson RL, Knowler WC, Sievers ML, Bennett PH, Looker
HC. Childhood obesity, other cardiovascular risk factors, and premature
death. N Engl J Med. 2010;362:485–493. doi: 10.1056/NEJMoa0904130.
16. Higgins JP, Whitehead A. Borrowing strength from external tri-
als in a meta-analysis. Stat Med. 1996;15:2733–2749. doi: 10.1002/
(SICI)1097-0258(19961230)15:24<2733::AID-SIM562>3.0.CO;2-0.
17. Psaty BM, Lumley T, Furberg CD, Schellenbaum G, Pahor M, Alderman
MH, Weiss NS. Health outcomes associated with various antihyperten-
sive therapies used as first-line agents: a network meta-analysis. JAMA.
2003;289:2534–2544. doi: 10.1001/jama.289.19.2534.
18. Bangalore S, Toklu B, Gianos E, Schwartzbard A, Weintraub H,
Ogedegbe G, Messerli FH. Optimal systolic blood pressure target
after SPRINT: insights from a network meta-analysis of randomized
trials. Am J Med. 2017;130:707.e8–719.e8. doi: 10.1016/j.amjmed.
2017.01.004.
19. Dias S, Sutton AJ, Welton NJ, Ades AE. Evidence synthesis for decision
making 3: heterogeneity–subgroups, meta-regression, bias, and bias-
adjustment. Med Decis Making. 2013;33:618−640.
20. Cipriani A, Furukawa TA, Salanti G, et al. Comparative efficacy and accept-
ability of 21 antidepressant drugs for the acute treatment of adults with
major depressive disorder: a systematic review and network meta-analy-
sis. Lancet. 2018;391:1357–1366. doi: 10.1016/S0140-6736(17)32802-7.
21. Brunton L, Parker K, Blumenthal D, Buxton I. Goodman and Gilman's
the Pharmacological Basis of Therapeutics, 2008. New York: McGraw-
Hill Education.
22. Meyers KE, Lieberman K, Solar-Yohay S, Han G, Shi V. The efficacy
and safety of valsartan in obese and non-obese pediatric hyperten-
sive patients. J Clin Hypertens (Greenwich). 2011;13:758–766. doi:
10.1111/j.1751-7176.2011.00502.x.
23. Wells T, Blumer J, Meyers KE, Neto JP, Meneses R, Litwin M, Vande
Walle J, Solar-Yohay S, Shi V, Han G; Valsartan Pediatric Hypertension
Study Group. Effectiveness and safety of valsartan in children aged 6 to
16 years with hypertension. J Clin Hypertens (Greenwich). 2011;13:357–
365. doi: 10.1111/j.1751-7176.2011.00432.x.
24. Li JS, Flynn JT, Portman R, Davis I, Ogawa M, Shi H, Pressler ML. The
efficacy and safety of the novel aldosterone antagonist eplerenone in chil-
dren with hypertension: a randomized, double-blind, dose-response study.
J Pediatr. 2010;157:282–287. doi: 10.1016/j.jpeds.2010.02.042.
25. Hazan L, Hernández Rodriguez OA, Bhorat AE, Miyazaki K, Tao
B, Heyrman R; Assessment of Efficacy and Safety of Olmesartan in
Pediatric Hypertension Study Group. A double-blind, dose-response
study of the efficacy and safety of olmesartan medoxomil in children and
adolescents with hypertension. Hypertension. 2010;55:1323–1330. doi:
10.1161/HYPERTENSIONAHA.109.147702.
 Burrello et al   Hypertension Management in Children   313
26. Wells TG, Portman R, Norman P, Haertter S, Davidai G, Fei Wang.
Safety, efficacy, and pharmacokinetics of telmisartan in pediatric
patients with hypertension. Clin Pediatr (Phila). 2010;49:938–946. doi:
10.1177/0009922810363609.
27. Batisky DL, Sorof JM, Sugg J, Llewellyn M, Klibaner M, Hainer JW,
Portman RJ, Falkner B; Toprol-XL Pediatric Hypertension Investigators.
Efficacy and safety of extended release metoprolol succinate in hyperten-
sive children 6 to 16 years of age: a clinical trial experience. J Pediatr.
2007;150:134–139, 139.e1. doi: 10.1016/j.jpeds.2006.09.034.
28. Shahinfar S, Cano F, Soffer BA, Ahmed T, Santoro EP, Zhang Z, Gleim G,
Miller K, Vogt B, Blumer J, Briazgounov I. A double-blind, dose-response
study of losartan in hypertensive children. Am J Hypertens. 2005;18(2 pt
1):183–190. doi: 10.1016/j.amjhyper.2004.09.009.
29. Flynn JT, Newburger JW, Daniels SR, Sanders SP, Portman RJ, Hogg RJ,
Saul JP; PATH-1 Investigators. A randomized, placebo-controlled trial of
amlodipine in children with hypertension. J Pediatr. 2004;145:353–359.
doi: 10.1016/j.jpeds.2004.04.009.
30. Li JS, Berezny K, Kilaru R, Hazan L, Portman R, Hogg R, Jenkins RD,
Kanani P, Cottrill CM, Mattoo TK, Zharkova L, Kozlova L, Weisman
I, Deitchman D, Califf RM. Is the extrapolated adult dose of fosino-
pril safe and effective in treating hypertensive children? Hypertension.
2004;44:289–293. doi: 10.1161/01.HYP.0000138069.68413.f0.
31. Soffer B, Zhang Z, Miller K, Vogt BA, Shahinfar S. A double-blind,
placebo-controlled, dose-response study of the effectiveness and safety
of lisinopril for children with hypertension. Am J Hypertens. 2003;16:
795–800.
32. Trachtman H, Frank R, Mahan JD, Portman R, Restaino I, Matoo TK,
Tou C, Klibaner M. Clinical trial of extended-release felodipine in pedi-
atric essential hypertension. Pediatr Nephrol. 2003;18:548–553. doi:
10.1007/s00467-003-1134-0.
33. Sorof JM, Cargo P, Graepel J, Humphrey D, King E, Rolf C, Cunningham
RJ. Beta-blocker/thiazide combination for treatment of hypertensive
children: a randomized double-blind, placebo-controlled trial. Pediatr
Nephrol. 2002;17:345–350. doi: 10.1007/s00467-002-0851-0.
34. Wells T, Frame V, Soffer B, Shaw W, Zhang Z, Herrera P, Shahinfar S;
Enalapril Pediatric Hypertension Collaborative Study Group. A double-blind,
placebo-controlled, dose-response study of the effectiveness and safety of
Downloaded from http://ahajournals.org by on July 31, 2018
enalapril for children with hypertension. J Clin Pharmacol. 2002;42:870–880.
35. Lu G, Ades AE. Combination of direct and indirect evidence in
mixed treatment comparisons. Stat Med. 2004;23:3105–3124. doi:
10.1002/sim.1875.
36. Welton NJ, Sutton AJ, Cooper N, Abrams KR, Ades AE. Evidence
Synthesis for Decision Making in Healthcare. Hoboken, NJ: John Wiley
& Sons, Ltd; 2012.
37. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis
JP, Clarke M, Devereaux PJ, Kleijnen J, Moher D. The PRISMA state-
ment for reporting systematic reviews and meta-analyses of studies that
evaluate healthcare interventions: explanation and elaboration. BMJ.
2009;339:b2700.
38. Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF. Improving
the quality of reports of meta-analyses of randomised controlled trials:
the QUOROM statement. Quality of reporting of meta-analyses. Lancet.
1999;354:1896–1900.
Novelty and Significance
What Is New?
• For the first time, the efficacy of different antihypertensive medications
in blood pressure reduction was compared head-to-head in pediatric
hypertensive patients through a network meta-analysis and meta-re-
gression approach.
What Is Relevant?
• Results from this study, including 13 studies for 2378 pediatric patients
affected by arterial hypertension indicate that angiotensin-converting
enzyme inhibitors (in particular lisinopril and enalapril) and angiotensin
receptor blockers (in particular losartan) are associated with a greater
systolic blood pressure and diastolic blood pressure reduction com-
pared with placebo, independently of the considered confounding fac-
39. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D,
Moher D, Becker BJ, Sipe TA, Thacker SB. Meta-analysis of observa-
tional studies in epidemiology: a proposal for reporting. Meta-Analysis
of Observational Studies in Epidemiology (MOOSE) group. JAMA.
2000;283:2008–2012.
40. Chen X, Wang Y. Tracking of blood pressure from childhood to adult-
hood: a systematic review and meta-regression analysis. Circulation.
2008;117:3171–3180. doi: 10.1161/CIRCULATIONAHA.107.730366.
41. Juhola J, Oikonen M, Magnussen CG, et al. Childhood physical, envi-
ronmental, and genetic predictors of adult hypertension: the cardiovas-
cular risk in young Finns study. Circulation. 2012;126:402–409. doi:
10.1161/CIRCULATIONAHA.111.085977.
42. Tirosh A, Afek A, Rudich A, Percik R, Gordon B, Ayalon N, Derazne
E, Tzur D, Gershnabel D, Grossman E, Karasik A, Shamiss A, Shai
I. Progression of normotensive adolescents to hypertensive adults:
a study of 26,980 teenagers. Hypertension. 2010;56:203–209. doi:
10.1161/HYPERTENSIONAHA.109.146415.
43. Welch WP, Yang W, Taylor-Zapata P, Flynn JT. Antihypertensive
drug use by children: are the drugs labeled and indi-
cated? J Clin Hypertens (Greenwich). 2012;14:388–395. doi:
10.1111/j.1751-7176.2012.00656.x.
44. Blowey DL. Update on the pharmacologic treatment of hypertension
in pediatrics. J Clin Hypertens (Greenwich). 2012;14:383–387. doi:
10.1111/j.1751-7176.2012.00659.x.
45. Chu PY, Campbell MJ, Miller SG, Hill KD. Anti-hypertensive drugs
in children and adolescents. World J Cardiol. 2014;6:234–244. doi:
10.4330/wjc.v6.i5.234.
46. Winnick S, Lucas DO, Hartman AL, Toll D. How do you improve compli-
ance? Pediatrics. 2005;115:e718–e724. doi: 10.1542/peds.2004-1133.
47. Giussani G, Bianchi E, Canelli V, Erba G, Franchi C, Nobili A, Sander JW,
Beghi E; EPIRES Group. Antiepileptic drug discontinuation by people
with epilepsy in the general population. Epilepsia. 2017;58:1524–1532.
doi: 10.1111/epi.13853.
48. Schmid CH, Stark PC, Berlin JA, Landais P, Lau J. Meta-regression
detected associations between heterogeneous treatment effects and study-
level, but not patient-level, factors. J Clin Epidemiol. 2004;57:683–697.
doi: 10.1016/j.jclinepi.2003.12.001.
49. Jansen JP. Network meta-analysis of individual and aggregate level data.
Res Synth Methods. 2012;3:177–190. doi: 10.1002/jrsm.1048.
50. Jansen JP, Cope S. Meta-regression models to address heterogeneity and
inconsistency in network meta-analysis of survival outcomes. BMC Med
Res Methodol. 2012;12:152. doi: 10.1186/1471-2288-12-152.
51. Dias S, Sutton AJ, Welton NJ, Ades AE. Heterogeneity: Subgroups, Meta-
Regression, Bias and Bias-Adjustment: NICE DSU Technical Support
Document No. 3. National Institute for Health and Clinical Excellence
[website]; 2011, last updated April 2012. http://www.nicedsu.org.uk.
Accessed May, 2018.
52. Ketteler M, Block GA, Evenepoel P, Fukagawa M, Herzog CA, McCann
L, Moe SM, Shroff R, Tonelli MA, Toussaint ND, Vervloet MG,
Leonard MB. Executive summary of the 2017 KDIGO Chronic Kidney
Disease-Mineral and Bone Disorder (CKD-MBD) Guideline Update:
what’s changed and why it matters. Kidney Int. 2017;92:26–36. doi:
10.1016/j.kint.2017.04.006.
tors (study-level summary measures of age, sex, baseline weight, and
baseline blood pressure).
Summary
Strong level of evidence is lacking to recommend any class of
antihypertensive medications over the others and the choice of a
particular drug is usually driven by hypertension underlying cause,
concurrent disorders, side effects, and clinician’s preference. Our
study suggests that angiotensin-converting enzyme inhibitors and
angiotensin receptor blockers could be the first choice in the treat-
ment of pediatric hypertension, even if and further head-to-head
randomized controlled trials are required to assess comparative
benefits and safety profile of each drug.