Vizza et al.

BMC Pulmonary Medicine (2017) 17:44

DOI 10.1186/s12890-017-0374-x

RESEARCH ARTICLE Open Access

Efficacy of 1, 5, and 20 mg oral sildenafil in

the treatment of adults with pulmonary
arterial hypertension: a randomized,
double-blind study with open-label
extension
Carmine Dario Vizza1* , B. K. S. Sastry2, Zeenat Safdar3, Lutz Harnisch4, Xiang Gao5, Min Zhang6,
Manisha Lamba5 and Zhi-Cheng Jing7

Abstract
Background: In a previous study, 6-minute walk distance (6MWD) improvement with sildenafil was not dose dependent
at the 3 doses tested (20, 40, and 80 mg 3 times daily [TID]). This study assessed whether lower doses were less effective
than the approved 20-mg TID dosage.
Methods: Treatment-naive patients with pulmonary arterial hypertension were randomized to 12 weeks of double-blind
sildenafil 1, 5, or 20 mg TID; 12 weeks of open-label sildenafil 20 mg TID followed. Changes from baseline in 6-minute
walk distance (6MWD) for sildenafil 1 or 5 mg versus 20 mg TID were compared using a Williams test. Hemodynamics,
functional class, and biomarkers were assessed.
Results: The study was prematurely terminated for non-safety reasons, with 129 of 219 planned patients treated. At week
12, 6MWD change from baseline was significantly greater for sildenafil 20 versus 1 mg (P = 0.011) but not versus 5 mg. At
week 24, 6MWD increases from baseline were larger in those initially randomized to 20 versus 5 or 1 mg (74 vs 50 and
47 m, respectively). At week 12, changes in hemodynamic parameters were generally small and variable between
treatment groups; odds ratios for improvement in functional class were not statistically significantly different.
Improvements in B-type natriuretic peptide levels were significantly greater with sildenafil 20 versus 1 but not 5 mg.
Conclusions: Sildenafil 20 mg TID appeared to be more effective than 1 mg TID for improving 6MWD; sildenafil 5 mg
TID appeared to have similar clinical and hemodynamic effects as 20 mg TID.
Trial registration: ClinicalTrials.gov NCT00430716 (Registration date: January 31, 2007).
Keywords: Sildenafil, Clinical trial, Pulmonary hypertension, Exercise test, Echocardiography, Dose

Background In the 12-week, randomized, double-blind, SUPER-1

Pulmonary arterial hypertension (PAH) is a fatal disease in study [3], statistically significant improvements in 6-minute
which increasing pulmonary vascular resistance ultimately walk distance (6MWD) were observed with sildenafil versus
culminates in right ventricular failure and death [1, 2]. placebo in treatment-naive patients at all 3 tested doses
The phosphodiesterase type 5 (PDE5) inhibitor sildenafil (20, 40, and 80 mg 3 times daily [TID]); improvements
is approved to treat adult patients with PAH [2]; pediatric were similar among groups and did not appear to be dose
use is approved in the European Union. related. However, hemodynamic parameters, including
mean pulmonary arterial pressure (mPAP), cardiac index,
* Correspondence: [email protected]; [email protected] and pulmonary vascular resistance index (PVRI), appeared
1
Department of Cardiovascular and Respiratory Disease, University of Rome
La Sapienza, Viale del Policlinico 155, 00161 Rome, Italy
to improve dose dependently with sildenafil treatment.
Full list of author information is available at the end of the article Sildenafil 20 mg TID appeared to reach the plateau of the
The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Vizza et al. BMC Pulmonary Medicine (2017) 17:44 Page 2 of 12

dose-response curve for 6MWD [3] and was confirmed by sildenafil. The hypothesis was that there is a dose that is
subsequent population pharmacokinetic and pharmacody- significantly less effective than sildenafil 20 mg TID.
namic analysis [4]. Secondary objectives included assessment of the safety
This study was conducted to fulfill a postapproval com- and tolerability of low-dose sildenafil during the 12 weeks
mitment from the US Food and Drug Administration of treatment in patients with PAH and evaluation of the
(FDA) to further explore the sildenafil dose-response effects of sildenafil on perceived PAH-progression bio-
curve. This multinational, randomized, double-blind study markers (B-type natriuretic peptide [BNP]/pro-BNP
investigated whether low doses of sildenafil (1 and 5 mg levels and tricuspid annular plane systolic excursion
TID) were less effective in adult patients with PAH than [TAPSE]). The study protocol and amendments were
the currently approved 20-mg TID dose. reviewed and approved by the Institutional Review
However, before completion of this low-dose study, Board and/or Independent Ethics Committee at each
results from another randomized, double-blind, placebo- participating center (Additional file 1); informed consent
controlled study (PACES-1) became available that sup- was obtained from all patients.
ported approval of a clinical worsening indication by the
FDA [5]. PACES-1 evaluated oral sildenafil in patients with Patients
PAH who were receiving stable epoprostenol therapy [6]. Patients were aged >18 years with idiopathic or heritable
In PACES-1, 75% of patients were titrated from sildenafil PAH or PAH associated with connective tissue disease
20 mg TID, received during the first 4 weeks, to sildenafil or surgical repair (5 years before enrollment) of atrial
40 mg TID at week 4, and then to sildenafil 80 mg TID at septal defect, ventricular septal defect, patent ductus
week 8 (and were maintained on this dose, as patients toler- arteriosus, or aorto-pulmonary window and 6MWD 100
ated). After 16 weeks, 6MWD, hemodynamic parameters, to 450 m. PAH, defined as mPAP 25 mmHg and pul-
and functional class improved. There was a significant delay monary artery wedge pressure 15 mmHg at rest (or a
in time to clinical worsening (TTCW) [6], defined as death, left ventricular end diastolic pressure <14 mmHg and
lung transplantation, hospitalization due to PAH, initiation absence of mitral stenosis on echocardiography), was
of bosentan therapy, or clinical deterioration requiring confirmed by right heart catheterization (RHC) within
a change in epoprostenol therapy, with sildenafil com- 12 weeks before randomization. Patients had to be on
pared with placebo. The effect was apparent by week 4, stable (30 days before RHC) doses of background
when all patients were receiving sildenafil 20 mg TID medication.
(P = 0.0074) [4]. Patients were excluded for use of PAH-specific ther-
Following approval of the clinical worsening indication apy, including prostacyclin, PDE5 inhibitors, and
in the United States in 2009, the FDA released Pfizer from endothelin-receptor antagonists (ETRAs); nitrates or ni-
the postapproval commitment to conduct a low-dose tric oxide donors; protease inhibitors, such as ritonavir
study. The study was subsequently terminated (June 2010) and saquinavir; ketoconazole, itraconazole, or other
based on the recommendation of the data monitoring strong cytochrome P450 (CYP) 3A4 inhibitors; and
committee (DMC) because sildenafil 20 mg TID had been alpha blockers. Patients previously receiving any of these
shown to reduce time to clinical worsening in PACES-1 drugs must have stopped use for 1 month before
and also acknowledging that with recruitment issues the screening. Concomitant medications were to remain
study was unlikely to meet original enrollment targets. stable throughout the treatment phase of the study;
Accumulated results are presented here. patients withdrew if they required additional PAH-
specific therapy.
Methods
Study design Assessments
Patients were stratified by baseline 6MWD (<325 or Six-minute walk distance was assessed at baseline (day
325 m) and PAH etiology and randomly assigned 1:1:1 1) and at weeks 4, 8, 12, 16, 20, and 24 as close to
to receive 12 weeks of treatment with sildenafil 1, 5, or sildenafil trough levels as possible (ie, just before dosing
20 mg TID, respectively, during the double-blind phase and 4 h after the last scheduled dose). Borg dyspnea
of the study (Fig. 1). Patients who completed the double- score was assessed at the end of the 6MWD evaluation.
blind phase were eligible for a 12-week, open-label Hemodynamic status was assessed at baseline and week
extension in which they received sildenafil 20 mg TID. 12, using RHC. World Health Organization functional
Patients who withdrew during the study were to be class was assessed at baseline; weeks 4, 8, 12, and 24;
followed up for safety assessments 30 days after the last and follow-up.
treatment date. Time to clinical worsening was assessed during the
The primary objective of the study was to demonstrate double-blind phase. Clinical worsening was defined as
a dose response for 6MWD for 1, 5, and 20 mg TID oral death, lung transplantation, hospitalization attributable
Vizza et al. BMC Pulmonary Medicine (2017) 17:44 Page 3 of 12

Double-Blind Phase (Dose Response) Open-Label Phase

Sildenafil 1 mg TID

Sildenafil 5 mg TID Sildenafil 20 mg TID

Screen Follow-up
Sildenafil 20 mg TID
End of double-
Randomization blind phase &
start of open-
label phase
21 days 30 days

1 4 8 12 16 20 24

Time, wk
Fig. 1 Study design. Legend: TID = 3 times daily

to pulmonary hypertension, or initiation of prostacyclin for 1 mg TID, sildenafil concentrations were anticipated
or ETRA therapy. to be at approximately EC50.
Blood samples for determination of BNP/pro-BNP
levels were collected at baseline and at weeks 1, 4, 8, 12, Pharmacokinetic modeling
16, 20, and 24. Echocardiography for TAPSE was per- Population modeling characterized sildenafil pharmaco-
formed at baseline and at weeks 4, 8, 12, and 24. A 2- kinetics; available sildenafil concentrations from all
dimensional Doppler examination was performed using patients across all visits were merged to develop a non-
an apical 4-chamber view. TAPSE index was measured linear mixed effects model (NONMEM, version 7.2;
as the total displacement of the tricuspid annulus (cm) ICON Development Solutions, Ellicott City, MD). Esti-
from end diastole to end systole, with values represent- mation was performed for underlying pharmacokinetic
ing the average TAPSE of 3 to 5 beats. parameters affecting the concentration-time profile.
For pharmacokinetic analysis, blood samples were col- Only covariates that were previously reported to affect
lected at the baseline visit (between 15 min and 3 h, >3 pharmacokinetic parameters [5] were tested in the
and 6 h, and >6 and 8 h postdose), week 1 (immediately model. To test for appropriateness, a visual predictive
after BNP/pro-BNP sampling), weeks 4 and 8 (immedi- check was performed by calculating the median and 90%
ately before 6MWD), and week 12 (between 15 min and prediction interval from 500 simulations of the resulting
3 h and between >3 and 6 h postdose, immediately be- population pharmacokinetic model.
fore 6MWD, between >6 and 8 h postdose, and during
RHC assessment). Statistical analysis
Adverse events (AEs) were monitored throughout the The estimated sample size was based on the primary end-
study. Laboratory testing and physical examinations were point and was determined using simulations. Assuming a
performed at screening, baseline, and weeks 4, 8, and 12. treatment effect of 30 m for sildenafil 20 versus 1 mg TID,
with a standard deviation of 60 m [3], 70 patients per
Dose selection group were required to detect a difference between treat-
The relationship between 6MWD and sildenafil expos- ments with 80% power at a 1-sided significance level of
ure could not be modeled because 6MWD had reached 2.5%. Allowing for 4% postrandomization nonevaluability,
a plateau across all SUPER-1 dose groups [3]. Therefore, approximately 219 patients (73 per group) were required
the relationship between PVRI and exposure was used to be randomized.
to select doses predicting exposures from the population For the primary endpoint, statistical significance was
pharmacokinetic/pharmacodynamic model. The average assessed with a 1-sided Williams trend test on the
sildenafil plasma concentration required to achieve 50% intent-to-treat (ITT) population; the ITT population
effect (EC50) on PVRI was approximately 3 ng/mL; at a consisted of randomized patients who received 1 dose
20-ng/mL concentration, sildenafil appeared to have a of study medication. The highest noneffective dose (ie,
90% maximal effect (EC90) on PVRI [4]. Therefore, after the highest dose that is statistically significantly different
receipt of 20 mg TID, sildenafil concentrations were an- from sildenafil 20 mg) was determined. Missing values
ticipated to be > EC90 for the entire 8-hour dosing inter- were replaced according to the last observation carried
val; for 5 mg TID, above EC50 for the entire 8-hour forward (LOCF) in the primary analysis and via multiple
dosing interval but < EC90 for most of the interval; and imputation for sensitivity analyses.
Vizza et al. BMC Pulmonary Medicine (2017) 17:44 Page 4 of 12

Additionally, changes in the primary endpoint were Figure 3 represents the sildenafil concentration data.
modeled by analysis of covariance (ANCOVA) with ran- Because a small accumulation existed between the first (at
domized treatment, baseline 6MWD, and etiology as baseline visit) and subsequent doses, only concentrations
stratification factors. Pairwise treatment group differ- after the second and subsequent doses are shown (for data
ences were estimated. In the open-label phase, changes including baseline visit, see Additional file 3). Concentra-
to week 24 (LOCF) were analyzed using this ANCOVA tions after concomitant administration of CYP3A4 inhibi-
model (but also including week 12 [LOCF] in the model) tors were adjusted for the estimated effect. Visual
if there was a nonmissing postweek-12 assessment. inspection of observed concentration distribution across
Secondary endpoints (including hemodynamic parame- each dose indicated consistency of the observed data with
ters) were assessed in the ITT population using LOCF; co- the model. In particular, in the 8 h after drug administra-
variates for each analysis included baseline value as well as tion, most of the determinations in the 1-mg TID group
the randomization strata of baseline 6MWD and etiology. had a concentration below 3 ng/mL, whereas in the 5-mg
Methods for LOCF, time to clinical worsening (TTCW), TID and 20-mg TID groups, most of the determinations
and Borg assessments are described in Additional file 2. For had a concentration above 3 ng/mL, which is the average
secondary endpoints, statistical significance was assessed sildenafil plasma concentration required to achieve 50%
based on nominal P values (<0.05; 2-sided) without adjust- effect (EC50) on PVRI [4]. An exploratory assessment (see
ment for multiplicity. Additional files 4 and 5) of the relationship between
6MWD, PVR, and steady-state concentrations revealed a
Results significant relationship for 6MWD, whereas only a small
The study was conducted at 34 centers in Europe, Asia, trend could be seen for PVR across the concentration
Russia, the United States, and Brazil. Of the planned 219 range observed (Additional files 6 and 7).
patients, 169 were screened, 130 were randomized, and 129
were treated (1 patient [sildenafil 1 mg] did not meet entry Six-minute walk distance
criteria). Treated patients were mostly female and mostly At week 12, compared with baseline, the increase in
Asian; baseline cardiac index was significantly higher in the 6MWD was of a magnitude consistent with estimates of
sildenafil 20-mg group versus the 1- and 5-mg groups clinical significance [7, 8] in 5- and 20-mg TID groups
(P = 0.0328 and 0.0030, respectively; Table 1). and smaller although statistically significant in the 1-mg
Patient disposition is shown in Fig. 2. Two patients TID group. Among dose groups, the mean change in
died during the double-blind phase (pneumonia [1 mg 6MWD from baseline was statistically significantly differ-
TID; death was the reason for discontinuation] and ent only for the sildenafil 20- versus 1-mg group (Fig. 4a).
acute exacerbation of idiopathic pulmonary fibrosis Analysis of change from baseline in 6MWD at week 12
[5 mg TID; patient was enrolled in error and received showed a statistically significant (P = 0.011) difference be-
4 days of treatment]), neither of which was considered tween sildenafil 1 mg and 20 mg, but not sildenafil 5 mg
to be treatment related; no deaths were reported in the and 20 mg (Table 2). The results were confirmed by an
open-label phase (Fig. 2). analysis of variance; the mean treatment difference
between sildenafil 20 mg and 1 mg was 23 (343) m and
Sildenafil concentration between 20 mg and 5 mg was 3 (23 to 17) m (P = 0.02
Overall, 129 patients provided 1068 sildenafil concentra- and 0.76, respectively).
tions. A 1-compartment pharmacokinetic model ad- Patients with baseline 6MWD <325 m at baseline had
equately described the sparse data. From this model, the greater increases in 6MWD after sildenafil treatment
estimated apparent clearance was 43.9 (95% CI, 39.3 than patients with baseline 6MWD 325 m (Fig. 4b).
48.6) L/h, the apparent volume of distribution was 458 Differences in 6MWD between Asian and non-Asian
(95% CI, 393523) L, and the absorption rate constant patients were noted for sildenafil 1 mg but not for 5 mg
was 2.16 (95% CI, 1.482.84) h1. Coadministration of or 20 mg (Fig. 5a and b); the number of non-Asian
weak or moderate CYP3A4 inhibitors (n = 12 patients/ patients was small.
110 samples) reduced CL/F by 40.4% (95% CI, 19.2% During the open-label period (weeks 12 to 24), in
61.6%). The model supported dose proportionality of which all patients received sildenafil 20 mg TID, patients
exposures. who received sildenafil 1 mg TID during the double-
The limit of quantification of the pharmacokinetic assay blind phase (weeks 0 to 12) had a larger increase in
was 1 ng/mL; 134 samples were below the limit of quanti- 6MWD than patients who received sildenafil 5 mg TID
fication (BLQ). The majority of BLQ samples (approxi- (mean change, 31 vs 6 m, respectively); the magnitude of
mately 75%) were measured at the 1-mg TID sildenafil change was similar between patients who received
dose, but had little effect on the population pharmacoki- sildenafil 1 mg and 20 mg TID in the double-blind phase
netic parameter estimates. (mean change, 31 vs 26 m; Fig. 4a).
Vizza et al. BMC Pulmonary Medicine (2017) 17:44 Page 5 of 12

Table 1 Baseline Patient Demographic and Clinical Table 1 Baseline Patient Demographic and Clinical
Characteristics Characteristics (Continued)
Baseline Characteristic Sildenafil Dose, TID MVO2, %f 63.4 (10.5) 63.0 (9.6) 64.3 (14.5)
1 mg 5 mg 20 mg Range 4182 4277 3190
(n = 41) (n = 43) (n = 45)
TAPSE indexg 1.25 (0.62) 1.2 (0.71) 1.36 (0.83)
Women, n (%) 28 (68) 33 (77) 26 (58)
Range 0.12.6 0.12.5 0.22.8
Age, y 42.5 (16.5) 44.4 (17.4) 46.4 (17.7)
Borg dyspnea scoreh 2.9 (2.5) 3.1 (1.9) 2.8 (2.1)
Range 1877 1878 2088
Range 010 08 09
Race, n (%)
All values are presented as mean (SD) unless stated otherwise
White 11 (27) 11 (26) 14 (31) 6MWD 6-minute walk distance, BMI body mass index, bpm beats per minute,
Black 2 (5) 2 (5) 1 (2) CTD connective tissue disease, mPAP mean pulmonary arterial pressure, MVO2
mixed venous oxygen saturation, PVR pulmonary vascular resistance, RAP right
Asian 27 (66) 30 (70) 30 (67) atrial pressure, TAPSE tricuspid annular plane systolic excursion, TID 3 times
daily, WHO World Health Organization
Other 1 (2) 0 0 a
n = 2 and 3 patients missing a baseline assessment in sildenafil 5- and 20-mg
Height, cm 159.0 (11.3) 160.2 (10.7) 160.7 (8.7) groups, respectively
b
n = 33, 32, and 33 patients contributing data in sildenafil 1-, 5-, and 20-mg
Range 130.0181.6 129.0189.0 147.0181.0 groups, respectively
c
n = 33, 33, and 34 patients contributing data in sildenafil 1-, 5-, and 20-mg
Weight, kg 61.7 (17.0) 63.1 (19.7) 61.4 (15.7)
groups, respectively
d
Range 32.0117.0 26.5126.1 35.0100.0 n = 33, 33, and 32 patients contributing data in sildenafil 1-, 5-, and 20-mg
groups, respectively
BMI, kg/m2 24.3 (5.6) 24.3 (6.6) 23.8 (6.2) e
n = 33, 32, and 32 patients contributing data in sildenafil 1-, 5-, and 20-mg
groups, respectively
Range 15.635.8 13.342.1 15.638.6 f
n = 33, 28, and 31 patients contributing data in sildenafil 1-, 5-, and 20-mg
WHO functional class, n (%) groups, respectively
g
n = 40 for all sildenafil groups
I 0 1 (2.3) 3 (6.7) h
n = 41, 40, and 42 patients contributing data in sildenafil 1-, 5-, and 20-mg
II 25 (61.0) 22 (51.2) 27 (60.0) groups, respectively

III 16 (39.0) 16 (37.2) 13 (28.9) Secondary and tertiary evaluations

IV 0 1 (2.3) 0 Hemodynamics
Missing 0 3 (7.0) 2 (4.4) Compared with baseline, there was a trend toward re-
Etiology, n (%) duction in pulmonary vascular resistance (PVR) at
Idiopathic 30 (73) 31 (72) 34 (76)
week 12 in all groups; the mean reduction was statis-
tically significantly different from 0 only in the 20-mg
Mean duration (range) 1.1 (06.7) 0.7 (06.5) 0.9 (014.9)
since diagnosis, y TID group (ie, 95% CIs do not include 0). There were
no statistically significant differences among treatment
Associated with CTD 6 (15) 8 (19) 5 (11)
groups for change in PVR (Table 3). Changes at week
Mean duration (range) 0.6 (02.3) 0.4 (01.8) 0.4 (01.9)
since diagnosis, y
12 in the additional hemodynamic parameters were
generally small and variable between groups.
Associated with 5 (12) 4 (9) 6 (13)
surgical repair
Mean duration (range) 5.9 (0.314.2) 3.5 (07.3) 4.5 (015.7)
since diagnosis, y Functional class and clinical worsening
Most patients in each treatment group remained in
6MWD, ma 347.5 (67.3) 347.7 (73.4) 340.4 (76.3)
the same functional class from baseline to week 12;
Range 167.5441.5 109.0455.0 114.0429.0
the same was true through week 24 (Table 4). Odds
b
Heart rate, bpm 83.6 (17.2) 78.9 (16.4) 80.1 (15.0) ratios (ORs) showed no significant differences for
Range 48122 42113 53110 functional class between sildenafil 20 mg and the 5-
RAP, mmHg c
10.5 (5.1) 10.1 (6.1) 8.4 (4.7) mg (OR, 1.08 [95% CI, 0.353.32]; P = 0.897) or 1-mg
Range 4.020.0 2.023.0 2.027.0 (OR, 1.55 [95% CI, 0.507.78]; P = 0.448) dose at
c week 12. Similarly, there were no differences between
mPAP, mmHg 57.2 (21.9) 55.4 (19.7) 51.1 (21.4)
sildenafil 20 mg and the 5-mg (OR, 1.31 [95% CI,
Range 25.0110.0 26.3117.0 25.0106.0
0.424.05]; P = 0.639) or 1-mg (OR, 0.93 [95% CI,
2d
Cardiac index, L/min/m 2.1 (0.7) 2.3 (0.6) 2.8 (1.2) 0.302.91]; P = 0.899) dose at week 24. Four patients
Range 1.03.5 1.03.8 1.15.9 (sildenafil 1 mg and 5 mg, n = 1 each; sildenafil
PVR, Wood units e
15.7 (9.9) 13.2 (8.3) 11.7 (9.1) 20 mg, n = 2) reported events defined as clinical wors-
Range 343 348 235 ening (initiation of ETRA therapy [sildenafil 5-mg
patient] and hospitalization due to PAH [all others]).
Vizza et al. BMC Pulmonary Medicine (2017) 17:44 Page 6 of 12

Screened (n=169)

Randomly assigned (n=130)

SIL 1 mg TID (n=42) SIL 5 mg TID (n=43) SIL 20 mg TID (n=45)

Treated (n=41) Treated (n=43) Treated (n=45)

Double-Blind

Discontinued (n=5) Discontinued (n=5) Discontinued (n=6)

Death (n=1) Withdrew consent (n=1) Withdrew consent (n=1)
Withdrew consent (n=2) Sponsor terminated study (n=1) Sponsor terminated study (n=2)
Sponsor terminated study (n=1) Adverse event (n=1) Adverse event (n=3)
Adverse event (n=1) Unrelated to study drug (n=1) Unrelated to study drug (n=1)
Related to study drug (n=1)* Other (n=2) Related to study drug (n=2)

Completed double-blind; Completed double-blind; Completed double-blind;

entered open-label (n=36) entered open-label (n=38) entered open-label (n=39)
Open-Label

Discontinued (n=9) Discontinued (n=7) Discontinued (n=7)

Lost to follow-up (n=1) Withdrew consent (n=1) Lost to follow-up (n=1)
Protocol violation (n=1) Sponsor terminated study (n=5) Withdrew consent (n=1)
Sponsor terminated study (n=4) Other (n=1) Sponsor terminated study (n=4)
Other (n=3) Other (n=1)

Completed (n=27) Completed (n=31) Completed (n=32)

Fig. 2 Patient disposition. Legend: TID = 3 times daily; SIL = sildenafil.*Right ventricular failure. Drug hypersensitivity (n = 1) and rash (n = 1)

Neurohormones 5 mg (P = 0.009 and 0.414, respectively). At week 24,

Decreases from baseline in BNP occurred in all changes from baseline were not significantly different
groups at week 12; the response was dose related among groups.
(Fig. 6a). The sildenafil 20-mg group was statistically
significantly (P = 0.005) different from the 1-mg but Echocardiography
not the 5-mg group (P = 0.496). At week 24, changes There was a trend toward a mean increase in TAPSE in
from baseline for sildenafil 20 mg were not signifi- all groups, but there were no statistically significant dif-
cantly different among groups. ferences in mean TAPSE index among groups (mean
Pro-BNP decreases occurred in all groups at week 12 [95% CI] increases of 0.14 [0.020.26], 0.17 [0.060.28],
and were dose related (Fig. 6b). Differences were signifi- and 0.04 [0.08 to 0.16] cm for sildenafil 1, 5, and
cant when sildenafil 20 mg was compared with 1 but not 20 mg TID, respectively, at week 12 [LOCF] and 0.21

Fig. 3 Plot of observed plasma sildenafil concentrations vs time after sildenafil dosing. Legend: Plasma sildenafil concentrations (open circles), sildenafil
doses of 1 mg (left), 5 mg (middle), and 20 mg (right). Median (solid line) and 90% prediction intervals (dashed lines) from simulations are overlaid. Tick
marks on the horizontal time axis indicate concentration measures below the limit of quantification. The shaded area shows the concentration range
between 3 ng/mL and 20 ng/mL, which are the average sildenafil plasma concentrations required to achieve 50% effect (EC50) and 90% effect (EC90)
on PVRI, respectively. TID = 3 times daily
Vizza et al. BMC Pulmonary Medicine (2017) 17:44 Page 7 of 12

A 90
70 m

Mean (SE) Change From

Baseline in 6MWD, m
47 m 50 m
60 *
41 m 38 m Sildenafil 1 mg TID
Sildenafil 5 mg TID
30 Sildenafil 20 mg TID
14 m
*P=0.011, sildenafil 20 vs 1 mg
TID; Williams trend test.
n=41 n=40 n=42 n=34 n=36 n=36
0
Week 12 Week 24

B 90
Mean (SE) Change From
Baseline in 6MWD, m

56 m
48 m
60
37 m
25 m 29 m

30
9m

n=13 n=14 n=14 n=28 n=26 n=28

0
Baseline 6MWD Baseline 6MWD
<325 m
Fig. 4 Mean change from baseline in 6MWD. Legend: Mean (SE) overall change from double-blind baseline in 6MWD in double-blind (week 12) and
open-label (week 24) phases of the study (a), and change from baseline to week 12 in 6MWD by baseline 6MWD (b). All patients received sildenafil 20 mg
TID in the open-label phase of the study (weeks 1324). 6MWD = 6-minute walk distance; TID = 3 times daily

[0.060.37], 0.40 [0.190.61], and 0.15 [0.09 to 0.39] at [LOCF] and 1.10 [1.75 to 0.46], 1.07 [1.55 to 0.58],
week 24 [LOCF]). and 0.28 [0.75 to 0.20] at week 24 [LOCF]), with no sig-
nificant differences between sildenafil 1- and 5-mg TID
Borg dyspnea score groups compared with sildenafil 20 mg TID.
Borg dyspnea scores trended toward reduction in all
groups (mean [95% CI] changes of 0.28 [0.76 to 0.20],
0.89 [1.35 to 0.43], and 0.43 [0.94 to 0.08] for sil- Correlations among parameters
denafil 1, 5, and 20 mg TID, respectively, at week 12 Baseline 6MWD was weakly correlated with BNP (r = 0.19;
P = 0.0393) and pro-BNP (r = 0.22; P = 0.0145). The change
in 6MWD at week 12 was also weakly correlated with
Table 2 Change From Baselinea in 6MWD at Week 12 (LOCF)
changes at week 12 in BNP (r = 0.18; P = 0.0499) and pro-
Williams Trend Test
BNP (r = 0.22; P = 0.0193).
Value Sildenafil Dose, TID
1 mg 5 mg 20 mg
(n = 41) (n = 43) (n = 45)
Adverse events
Least squares mean 14.21 40.75 38.36
The overall number of AEs and numbers of patients
MLE meanb 14.21 39.52 39.52
reporting AEs were similar between treatment groups
Mean differencec 24.15 1.17 in the double-blind and open-label portions of the
Williams statistic 2.37 0.11 study; treatment-related AEs (number of AEs and
97.5% lower confidence limit 3.37 21.48 patients reporting AEs) increased with increasing dose
P valued 0.011 0.545 (Table 5). Sildenafil was generally well tolerated, with
6MWD 6-minute walk distance, LOCF last observation carried forward, MLE
most AEs being mild or moderate in severity. Dyspnea
maximum likelihood estimation, TID 3 times daily was the most common AE reported in both phases of the
a
Baseline is the average of the screening and day 1 values study; headache was the most common treatment-related
b
MLE mean is defined as least squares mean if it satisfies descending response
relationship for descending doses; if descending relationship does not hold, AE (Table 5). No patients discontinued as a result of ab-
MLE mean is defined as weighted mean of adjacent least squares means
c
normal laboratory test results, and there was no evidence
Mean difference was calculated as the least squares mean for sildenafil 20 mg
minus the MLE mean for sildenafil lower dose
of dose-related increase in laboratory test abnormalities
d
From directional test vs 20 mg TID with increasing sildenafil dose.
Vizza et al. BMC Pulmonary Medicine (2017) 17:44 Page 8 of 12

A
90 Sildenafil 1 mg TID
Asian Non-Asian Sildenafil 5 mg TID
43 m Sildenafil 20 mg TID
48 m
Mean (SE) Change From
Baseline in 6MWD, m
60
42 m 41 m
29 m

30

n=22 n=26 n=23 9 m n=11 n=14

0
n=13

30

B Asian Non-Asian
90 55 m
Mean (SE) Change From

90
60 m
Baseline in 6MWD, m

59 m
52 m
36 m
42 m
60 60
31 m
38 m
28 m 32 m
9m
30 30

n=8 n=8 n=7 n=14 n=18 n=16 n=4 n=4 n=5 17 m n=7 n=9
0 0
n=9
Baseline 6MWD Baseline 6MWD
<325 m
Baseline 6MWD Baseline 6MWD
30 <325 m
Fig. 5 Mean change from baseline in 6MWD assessed by race. Legend: Mean (SE) overall change from double-blind baseline in 6MWD in the
double-blind (week 12) phase of the study (a) and change from baseline to week 12 in 6MWD by baseline 6MWD (b) assessed by race (Asian vs
non-Asian). 6MWD = 6-minute walk distance; TID = 3 times daily

Discussion could be seen as a placebo effect due to participation

Sildenafil is one of the most widely used drugs in the in an RCT.
treatment of PAH. The dose of 20 mg TID was approved Among dose groups, the change in 6MWD from base-
based on the results of the SUPER-1 study which dem- line was significant only with sildenafil 20 mg TID com-
onstrated that Sildenafil 20 mg TID appeared to reach pared with sildenafil 1 mg TID. A Williams trend test
the plateau of the dose-response curve for 6MWD, des- confirmed that sildenafil 1 mg TID was the only dose
pite the larger hemodynamic effects seen with the high- statistically inferior to the approved dose of 20 mg TID.
est dosage (80 mg TID). These results raise the question Generally, patients had greater improvements in
as to whether a lower dosage could have a similar effect hemodynamic parameters with sildenafil 20 mg TID ver-
on 6MWD compared to the approved dose. This aspect sus 1 mg TID; however, these improvements were not
was addressed in the present study. statistically significantly different. Significant differences
We found a significant increase from baseline in were observed between sildenafil 1 mg TID and 20 mg
6MWD at 12 weeks with all sildenafil doses; however, TID for neurohormones at week 12.
only at higher doses (5 and 20 mg TID) was the im- There were no statistically significant differences between
provement of a magnitude considered to be clinically sildenafil 20 and 5 mg TID in 6MWD, hemodynamics, or
relevant (~40 m) [7, 8]. In the absence of a placebo con- changes in functional class.
trol arm, the small non-clinically significant increase in Results from pharmacokinetic modeling showed that
6MWD in the 1 mg TID group in the double blind the observed exposure with sildenafil 1 mg TID was
phase should be interpreted with caution as being a slightly below EC50 for maximal PVR change, the ob-
treatment effect as it is possible that this improvement served exposure with sildenafil 5 mg TID was above
Vizza et al. BMC Pulmonary Medicine (2017) 17:44 Page 9 of 12

Table 3 Adjusted Change From Baseline in Hemodynamic Parameters at Week 12

Baseline Characteristic Sildenafil Dose, TID
1 mg 5 mg 20 mg
Heart rate
n 33 32 33
LS mean (95% CI), bpm 3.4 (1.1 to 7.9) 0.7 (5.2 to 3.7) 5.0 (9.3 to 0.8)
P value vs 20 mg TID 0.0019 0.1066
RAP
n 33 33 34
LS mean (95% CI), mmHg 0.5 (2.3 to 1.2) 0.8 (2.5 to 0.9) 1.7 (3.3 to 0)
P value vs 20 mg TID 0.2741 0.4098
mPAP
n 33 33 34
LS mean (95% CI), mmHg 0.1 (4.0 to 3.7) 2.2 (5.9 to 1.5) 2.6 (6.2 to 0.9)
P value vs 20 mg TID 0.2776 0.8458
Cardiac index
n 32 31 30
LS mean (95% CI), L/min/m2 0.1 (0.2 to 0.3) 0.1 (0.1 to 0.4) 0.1 (0.2 to 0.3)
P value vs 20 mg TID 0.9023 0.7590
PVR
n 32 31 30
LS mean (95% CI), Wood units 1.2 (3.3 to 0.9) 2.0 (4.1 to 0) 2.4 (4.3 to 0.4)
P value vs 20 mg TID 0.3694 0.8010
PVRI
n 32 31 30
LS mean (95% CI), Wood units*m2 1.7 (4.9 to 1.5) 3.1 (6.2 to 0) 3.5 (6.4 to 0.5)
P value vs 20 mg TID 0.3868 0.8628
MVO2
n 33 28 31
LS mean (95% CI), % 1.5 (2.2 to 5.2) 3.0 (0.8 to 6.7) 3.0 (0.4 to 6.4)
P value vs 20 mg TID 0.4918 0.9791
bpm beats per minute, LS least squares, mPAP mean pulmonary arterial pressure, MVO2 mixed venous oxygen saturation, PVR pulmonary vascular resistance, PVRI
PVR index, RAP right atrial pressure, TID 3 times daily

Table 4 Change From Baseline to Weeks 12 and 24 in Functional Class (LOCF)

Change, n (%) Sildenafil Dose, TID
Double-Blind Phase (Week 12) Open-Label Phase (Week 24)
1 mg (n = 41) 5 mg (n = 43) 20 mg (n = 45) 1 mg (n = 41) 5 mg (n = 43) 20 mg (n = 45)
Worsened 2 classes 0 0 0 0 0 1 (2)
Worsened 1 class 1 (2) 3 (7) 2 (4) 0 3 (7) 1 (2)
No change 35 (85) 27 (63) 35 (78) 23 (56) 19 (44) 22 (49)
Improved 1 class 4 (10) 10 (23) 6 (13) 11 (27) 13 (30) 11 (24)
Improved 2 classes 1 (2) 0 0 1 (2) 1 (2) 1 (2)
Missing 0 3 (7) 2 (4) 6 (15) 7 (16) 9 (20)
LOCF last observation carried forward, TID 3 times daily
Vizza et al. BMC Pulmonary Medicine (2017) 17:44 Page 10 of 12

EC50 and approaching EC90, and the observed exposure

A Sildenafil 1 mg TID
with sildenafil 20 mg TID was mainly above EC90.
Sildenafil 5 mg TID
Sildenafil 20 mg TID
The pharmacokinetic data justify the different clinical
responses between sildenafil 1 and 20 mg TID and explain
*P=0.005 and P=0.009,
Mean Change (95% CI) From

90 sildenafil 20 vs 1 mg TID. the small difference observed between sildenafil 20 and

Baseline in BNP, pg/mL

60 5 mg TID because most of the patients on 5 mg TID had

30 a sildenafil plasma level between 3 and 20 ng/mL.
0
n=38 n=39 n=42 n=29 n=33 n=34 A significant correlation among mean sildenafil plasma
30
60
concentration and 6MWD could be observed, although
90 the relationship between average sildenafil plasma concen-
120 tration and PVR appeared to show only a shallow trend.
150 Whether this was due to the missing placebo group or
180 * was a consequence of the smaller sample size and larger
Week 12 Week 24
variability on PVR cannot be concluded but should be
B interpreted on the basis of the complex interplay between
Mean Change (95% CI) From
Baseline in Pro-BNP, pg/mL

300 pharmacokinetics and pharmacodynamics. The vasodila-

tor effect is the result of the interplay of several factors:
0
n=39 n=40 n=43 n=30 n=33 n=34
tissue penetration of the drug, density and activity of
PDE5 enzyme, and severity of vascular lesions.
300 Smaller improvement in 6MWD at week 12 with silden-
afil 20 mg TID was observed in this study (38 m) com-
600 pared with SUPER-1 (45 m); however, patient populations
differed. Both studies had similar baseline 6MWD, but a
900 greater proportion of patients in this study had baseline
Week 12 Week 24
functional class II status compared with those in SUPER-1
Fig. 6 Changes from baseline in BNP (a) and pro-BNP (b) during (57% vs 39%, respectively); therefore, patients in this study
double-blind (week 12) and open-label (week 24) phases of the study. All
patients received sildenafil 20 mg TID in the open-label phase of the
had lower-than-expected 6MWD at baseline. Patients in
study (weeks 1324). BNP = B-type natriuretic peptide; TID = 3 times daily our study were also younger (45 vs 49 years), with a shorter
time since diagnosis (median, 0.17 vs 0.85 years) and an in-
creased percentage of Asian patients (67% vs 7%). Geo-
graphic variation in 6MWD has been described for

Table 5 Adverse Event Summary

All-Cause (Treatment-Related) Sildenafil Dose, TID
AEs, n
Double-Blind Phase (Week 12) Open-Label Phasea (Week 24)
1 mg (n = 41) 5 mg (n = 43) 20 mg (n = 45) 1 mg (n = 41) 5 mg (n = 43) 20 mg (n = 45)
Patients with AEs 17 (9) 17 (10) 19 (14) 23 (11) 22 (12) 22 (15)
Patients with serious AEs 4 (0) 2 (0) 3 (1) 6 (0) 3 (0) 5 (2)
Discontinuations due to AEs 1 (1) 1 (0) 3 (2) 1 (1) 1 (0) 3 (2)
Deaths 1 (0) 1 (0) 0 0 0 0
Number of AEs 46 (12) 41 (17) 47 (24) 90 (19) 69 (27) 74 (31)
AEs occurring in 3 patients
Anemia 1 (0) 0 (0) 3 (1) 1 (0) 1 (0) 3 (1)
Fatigue 2 (1) 1 (0) 0 (0) 3 (1) 1 (0) 0 (0)
Nasopharyngitis 2 (0) 1 (0) 1 (0) 2 (0) 1 (0) 3 (0)
Dizziness 2 (1) 1 (0) 1 (1) 3 (1) 2 (1) 2 (2)
Dyspnea 2 (0) 3 (0) 3 (0) 2 (0) 4 (1) 3 (0)
Headache 1 (1) 1 (1) 3 (3) 2 (2) 3 (2) 3 (3)
Epistaxis 0 (0) 2 (2) 0 (0) 1 (0) 3 (3) 0 (0)
Back pain 0 (0) 1 (0) 2 (1) 1 (0) 1 (0) 3 (2)
AE adverse event, TID 3 times daily
a
Includes AEs from the double-blind and open-label portions of the study
Vizza et al. BMC Pulmonary Medicine (2017) 17:44 Page 11 of 12

patients with PAH and was reported to be independent of study, similar results may have been observed. A nonin-
anthropometric factors [9]. Although few non-Asian feriority study comparing sildenafil 5 mg TID versus
patients enrolled in this study, 6MWD did not appear to 20 mg TID would require an unrealistically large sample
differ between groups, with the exception of sildenafil size for a rare disease like PAH. Estimating from the re-
1 mg TID (Fig. 5). sults of the current study, 382 patients would be required
Interestingly, results from the open-label phase suggest for a study with a noninferiority margin of 15 m at 90%
the possibility of further improvement in 6MWD after the power and a 1-sided significance level of 0.05, assuming a
first 3 months of therapy with sildenafil 20 mg TID. The true difference (5 vs 20 mg TID) of 0 m and a standard
mean increase in 6MWD from baseline at the end of the deviation of 50 m. The required sample size would in-
double-blind phase (41 m) was maintained in the sildenafil crease if patient dropout was considered or if a smaller
5-mg group uptitrated to sildenafil 20 mg TID in the ex- noninferiority margin was desired.
tension study (50 m), yet larger increases were observed
from the end of the double-blind study to the end of the
open-label study in the sildenafil 1- and 20-mg groups Conclusion
(from 1447 m and from 3870 m, respectively). Thus, Despite this study having the limitation of premature ter-
sildenafil 20 mg TID maintains treatment effects regard- mination, sildenafil 1 mg TID, but not 5 mg TID, was
less of prior low-dose treatment. However, 6MWD did not shown to be inferior to 20 mg TID for improvement in
increase to the same degree in patients previously treated 6MWD in patients with PAH. Sildenafil 5 mg TID ap-
with lower doses as in patients who continuously received peared to have similar clinical and hemodynamic effects as
20 mg TID, suggesting that a longer duration of an ad- 20 mg TID. Interestingly, 6MWD results from the open-
equate dose may confer a larger improvement in 6MWD. label phase of the study suggest that patients on the
Interestingly, the total improvement observed after approved sildenafil dose (20 mg TID) continued to show
24 weeks in the 20-mg group (70 m) was larger than in clinical improvement after the first 12 weeks of treatment.
the SUPER-1 study at 12 weeks (48 m) or 1 year (51 m) Hence, the question remains whether doses lower than
for all sildenafil doses combined. It may be possible that in 20 mg TID have therapeutic value and needs to be seen in
a population of young and mainly incident cases, as in our light of the current therapeutic approach in PAH.
study, further improvements in 6MWD may be observed
with continued sildenafil treatment.
Additional files
Decreases for BNP and pro-BNP versus baseline were
significantly higher with sildenafil 20 mg versus 1 mg Additional file 1: List of Investigators and Corresponding Ethics
TID at week 12, paralleling findings with 6MWD. BNP Committees or Institutional Review Boards. (PDF 140 kb)
levels similarly paralleled improvements (BNP levels de- Additional file 2: Supplemental Methods. Methods for LOCF, time to
creased) or worsening (BNP levels increased) in pulmon- clinical worsening (TTCW), and Borg assessments. (DOCX 14 kb)
ary hemodynamics and functional parameters, including Additional file 3: Figure S1. Relationship of change from baseline in
6MWD and sildenafil average steady-state concentration. (DOCX 121 kb)
6MWD, in patients with PAH in a previous study [10].
Additional file 4: Figure S2. Relationship of change from baseline in
Elevated plasma BNP levels are associated with in- PVR and sildenafil average steady-state concentration. (DOCX 125 kb)
creased mortality in patients with PAH, and a decrease Additional file 5: Figure S3. Plot of observed plasma sildenafil
in BNP levels after therapy is associated with improved concentrations (open circles) vs time after sildenafil doses of 1 mg TID
survival [11, 12]. Pro-BNP levels have recently been (bottom panels), 5 mg TID (middle panels), and 20 mg TID (top panels).
(DOCX 179 kb)
shown to identify poor outcome in patients with PAH
Additional file 6: Table S1. Results of a linear model, regressing 6MWD
[13, 14]. Longer-term follow-up of patients from our change from baseline against sildenafil average steady-state concentrations.
study is not ongoing, which prevents any correlation (DOCX 14 kb)
with mortality. Additional file 7: Table S2. Results of a linear model, regressing PVR
The main limitation of the present study is its prema- change from baseline against sildenafil average steady state concentrations.
(DOCX 14 kb)
ture termination. The study was designed to assess the
relative efficacy of sildenafil 20 mg TID and lower doses
and powered for the primary endpoint but the sample Abbreviations
size was not reached because of premature termination 6MWD: 6-minute walk distance; AE: Adverse event; ANCOVA: Analysis of
[4, 6]. Looking at the results, this does not seem a major covariance; BLQ: Below the limit of quantification; BNP: B-type natriuretic peptide;
CYP: Cytochrome P450; DMC: Data monitoring committee; ETRA: Endothelin-
issue, as the difference in the primary and secondary receptor antagonists; FDA: US Food and Drug Administration; ITT: Intent to treat;
endpoints between 1 mgTID and 20 mg TID is statisti- LOCF: Last observation carried forward; mPAP: Mean pulmonary arterial pressure;
cally significant and coherent. Regarding the comparison OR: Odds ratio; PAH: Pulmonary arterial hypertension; PDE5: Phosphodiesterase
type 5; PVR: Pulmonary vascular resistance; PVRI: Pulmonary vascular resistance
between the 5-mg and 20-mg groups, the differences index; RHC: Right heart catheterization; TAPSE: Tricuspid annular plane systolic
were small enough that, even with the completion of the excursion; TID: 3 times daily; TTCW: Time to clinical worsening
Vizza et al. BMC Pulmonary Medicine (2017) 17:44 Page 12 of 12

Acknowledgements 2. Galie N, Corris PA, Frost A, Girgis RE, Granton J, Jing ZC, et al. Updated
The authors would like to thank Susan Raber for assistance with the analyses treatment algorithm of pulmonary arterial hypertension. J Am Coll Cardiol.
presented in the manuscript. Editorial support was provided by Tiffany Brake, 2013;62:D6072.
PhD, Candace Lundin, DVM, MS, and Janet E. Matsuura, PhD, from Complete 3. Galie N, Ghofrani HA, Torbicki A, Barst RJ, Rubin LJ, Badesch D, et al.
Healthcare Communications, LLC, and was funded by Pfizer Inc. Sildenafil citrate therapy for pulmonary arterial hypertension. N Engl J Med.
2005;353:214857.
4. Harnisch L, Hayashi N. Population pharmacokinetic (PK) of sildenafil in
Funding
paediatric and adult pulmonary arterial hypertension (PAH) patients. Eur
This study was funded by Pfizer Inc.
Respir J 2009;34(Suppl. 53):3916.
5. Revatio (sildenafil citrate). Full Prescribing Information. New York: Pfizer Inc; 2014.
Availability of data and materials 6. Simonneau G, Rubin LJ, Galie N, Barst RJ, Fleming TR, Frost AE, et al.
Pfizers policies on the provision of clinical trial data are set out on our website: Addition of sildenafil to long-term intravenous epoprostenol therapy in
http://www.pfizer.com/research/clinical_trials/trial_data_and_results. In addition to patients with pulmonary arterial hypertension: a randomized trial. Ann
posting clinical trial results on the clinicaltrials.gov registry, Pfizer provides secure Intern Med. 2008;149:52130.
access to anonymized patient-level data to qualified researchers in response to 7. Gabler NB, French B, Strom BL, Palevsky HI, Taichman DB, Kawut SM, et al.
scientifically valid research proposals. Further detail can be found at: http:// Validation of 6-minute walk distance as a surrogate end point in pulmonary
www.pfizer.com/research/clinical_trials/trial_data_and_results/data_requests. arterial hypertension trials. Circulation. 2012;126:34956.
Pfizers practices adhere to the principles for responsible data sharing laid out by 8. Mathai SC, Puhan MA, Lam D, Wise RA. The minimal important difference in
the European Federation of Pharmaceutical Industries and Associations (EFPIA) the 6-minute walk test for patients with pulmonary arterial hypertension.
and the Pharmaceutical Research and Manufacturers of America (PhRMA): http:// Am J Respir Crit Care Med. 2012;186:42833.
phrma.org/sites/default/files/pdf/ 9. Casanova C, Celli BR, Barria P, Casas A, Cote C, de Torres JP, et al. The 6-min
PhRMAPrinciplesForResponsibleClinicalTrialDataSharing.pdf. walk distance in healthy subjects: reference standards from seven countries.
Eur Respir J. 2011;37:1506.
Authors contributions 10. Leuchte HH, Holzapfel M, Baumgartner RA, Neurohr C, Vogeser M, Behr J.
CDV, BKSS, ZS, LH, XG and Z-CJ participated in the acquisition of data, analysis and Characterization of brain natriuretic peptide in long-term follow-up of
interpretation of data, and drafted the manuscript or revised it critically for pulmonary arterial hypertension. Chest. 2005;128:236874.
intellectual content; MZ performed the statistical analysis, analyzed and interpreted 11. Casserly B, Klinger JR. Brain natriuretic peptide in pulmonary arterial
the data, and drafted the manuscript or revised it critically for intellectual content; hypertension: biomarker and potential therapeutic agent. Drug Des Devel
ML analyzed and interpreted the data and drafted the manuscript or revised it Ther. 2009;3:26987.
critically for intellectual content. All authors read and approved the final 12. Nagaya N, Nishikimi T, Uematsu M, Satoh T, Kyotani S, Sakamaki F, et al.
manuscript. Plasma brain natriuretic peptide as a prognostic indicator in patients with
primary pulmonary hypertension. Circulation. 2000;102:86570.
13. Mauritz GJ, Rizopoulos D, Groepenhoff H, Tiede H, Felix J, Eilers P, et al.
Competing interests Usefulness of serial N-terminal pro-B-type natriuretic peptide measurements
C.D.V. has received fees for serving as a speaker, consultant, and advisory for determining prognosis in patients with pulmonary arterial hypertension.
board member from Actelion, Domp, GlaxoSmithKline, Italfarmaco, Lilly, Am J Cardiol. 2011;108:164550.
Pfizer, and United Therapeutics. B.K.S.S. has received research funding from 14. Fijalkowska A, Kurzyna M, Torbicki A, Szewczyk G, Florczyk M, Pruszczyk P,
Pfizer and Actelion. Z.S. has been a consultant and served as a speaker and et al. Serum N-terminal brain natriuretic peptide as a prognostic parameter
advisory board member for United Therapeutics, Gilead, and Actelion. L.H., in patients with pulmonary hypertension. Chest. 2006;129:131321.
M.L., and M.Z. are Pfizer employees. X.G. is a former Pfizer employee. Z.-C.J.
has received fees for serving as a speaker, consultant, and advisory board
member from Actelion, Bayer, GlaxoSmithKline, Lilly, Pfizer, and United
Therapeutics.

Ethics approval and consent to participate

The study protocol and amendments were reviewed and approved by the
Institutional Review Board and/or Independent Ethics Committee at each
participating center (listed in Additional file 1); informed consent was
obtained from all patients.

Author details
1
Department of Cardiovascular and Respiratory Disease, University of Rome
La Sapienza, Viale del Policlinico 155, 00161 Rome, Italy. 2CARE Hospitals,
Gandhi Bhavan Road Nampally, Hyderabad, India. 3Baylor College of
Medicine, 1 Baylor Plaza, Houston, TX 77030, USA. 4Pfizer Ltd, Ramsgate
Road, Sandwich Kent CT13 9NJ, UK. 5Pfizer Inc, 558 Eastern Point Rd, Groton,
CT 06340, USA. 6Pfizer Inc, 10646 Science Center Dr, La Jolla Campus, San
Diego, CA 92121, USA. 7Shanghai Pulmonary Hospital, Tongji University Submit your next manuscript to BioMed Central
School of Medicine, 507, Zhengmin Road, Shanghai, China.
and we will help you at every step:
Received: 13 May 2016 Accepted: 31 January 2017 We accept pre-submission inquiries
Our selector tool helps you to find the most relevant journal
We provide round the clock customer support
References
1. McLaughlin VV, Archer SL, Badesch DB, Barst RJ, Farber HW, Lindner JR, et al. Convenient online submission
ACCF/AHA 2009 expert consensus document on pulmonary hypertension: a Thorough peer review
report of the American College of Cardiology Foundation Task Force on
Inclusion in PubMed and all major indexing services
Expert Consensus Documents and the American Heart Association developed
in collaboration with the American College of Chest Physicians; American Maximum visibility for your research
Thoracic Society, Inc.; and the Pulmonary Hypertension Association. J Am Coll
Cardiol. 2009;53:1573619. Submit your manuscript at
www.biomedcentral.com/submit