Content uploaded by Yi-Hsing Chen
Author content
All content in this area was uploaded by Yi-Hsing Chen on Jan 05, 2016
Content may be subject to copyright.
Nephrol Dial Transplant (2009) 24: 1946–1951
doi: 10.1093/ndt/gfn751
Advance Access publication 22 January 2009
Systemic inflammation is associated with pulmonary hypertension
in patients undergoing haemodialysis
Tung-Min Yu1,4, Yi-Hsing Chen2,4, Jeng-Yuan Hsu3,5, Chung-Shu Sun6, Ya-Wen Chuang1,4,
Cheng-Hsu Chen1,4, Ming-Ju Wu1,5, Chi-Hung Cheng1,4and Kuo-Hsiung Shu1,5
1Division of Nephrology, 2Division of Allergy, Immunology, 3Division of Chest Medicine, 4Department of Internal Medicine,
Taichung Veterans General Hospital, 5Chung-Shan Medical University and 6Division of Pediatry, Zhong-Xing branch, Taipei City
Hospital, Taiwan
Correspondence and offprint requests to: Kuo-Hsiung Shu; E-mail: khshu@vghtc.gov.tw
Abstract
Background. Pulmonary hypertension (PH) is an over-
looked cardiovascular morbidity in patients undergoing
haemodialysis. Inflammation has been demonstrated to
play a significant role with certain types of PH in non-
uraemic patients, but studies analysing the mechanisms
in dialyzed patients with PH are rare. Hence, we investi-
gated systemic and local inflammation biomarkers associ-
ated with PH in uraemia patients to elucidate the potential
mechanism.
Methods. A cross-sectional study was conducted in which
97 haemodialysis patients were initially evaluated in our
hospital. Twelve inflammatory cytokines were measured
using a cytometric beads assay in patients with and without
PH. FENO (fractional exhaled nitric oxide) was checked by
a chemiluminescence analyser in patients with and without
PH as well as by normal controls.
Results. Thirty-nine eligible patients were enrolled. Com-
pared to patients without PH (group A), patients with PH
(group B) had significantly higher serum levels of hs-
CRP, IL-1β,TNF-αand IL-6. FENO was also measured.
Though the pre-dialysis FENO levels were elevated in both
groups; group B patients had significantly higher pre-
dialysis FENO levels than group A patients (39.9 ±16.7
versus 31.8 ±10.3, P=0.045). The post-dialysis FENO lev-
els returned to normal in group A while the remaining
were significantly higher in group B (30.3 ±10.3 versus
20.1 ±10.9, P=0.003).
Conclusions. Our study revealed that dialyzed patients with
PH had a significantly higher level of airway FENO as well as
serum levels of acute phase reactive protein and cytokines,
including IL-1β,TNF-αand IL-6. A chronic inflammation
might play an important role in the pathogenesis of PH in
patients undergoing haemodialysis.
Keywords: fractional exhaled nitric oxide; haemodialysis; inflammatory
cytokines; pulmonary hypertension
Introduction
Pulmonary hypertension (PH) is a condition with high mor-
bidity and mortality. It is defined as an elevation of the mean
pulmonary arterial pressure (Ppa )≥35 mmHg and can be
the result of heart, lung or other systemic diseases [1].
The pathogenesis of PH involves endothelial and vascular
smooth cell dysfunction and obliteration of the lumen of
small vessels in the lungs known as plexogenic arteriopa-
thy, which results in increased vascular resistance to blood
flow. PH is also a common complication in various sys-
temic inflammatory conditions, and its resemblance to the
pathological picture of different underlying diseases sug-
gests an identical pathophysiology [2]. Tuder et al. were the
first to identify inflammatory infiltrates in the area of plex-
iform vascular lesions in primary PH [3] and the further
in vitro study demonstrated that inflammation contributes
to the growth of pulmonary plexiform lesions in various cir-
cumstances [4]. Recent experimental data have suggested
that several inflammatory cytokines play a crucial role in
the regulation of pulmonary artery pressure [5–7]. In pa-
tients with end-stage renal disease (ESRD), PH has been a
neglected morbidity. Yigla et al. found unexplained PH in
some long-term haemodialysis patients and reported a strik-
ingly high prevalence of 40% in uraemic patients, detected
by Doppler echocardiography [8,9]. The pathogenesis of
PH in patients with ESRD is complex and remains un-
determined. In addition to the detrimental haemodynamic
impact of arterio-venous fistula, which leads to a high car-
diac output in haemodialysis, we suggested that various
metabolic and inflammatory dearrangements in the uraemic
milieu might play a role in the pathogenesis of PH. The aim
of the study was to investigate the inflammatory status of
systemic and local pulmonary tissue by the measurement
of various plasma cytokines and FENO (fractional exhaled
nitric oxide), an inflammatory marker of airway and lung
parenchyma, in patients with and without PH undergoing
long-term haemodialysis.
CThe Author [2009]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
For Permissions, please e-mail: journals.permissions@oxfordjournals.org
by guest on November 3, 2015http://ndt.oxfordjournals.org/Downloaded from
Systemic inflammation is associated with PH 1947
97 haemodialysis patients were evaluated initially
47 patients were excluded due to past history of
coronary artery disease ( n = 25), autoimmune
disease (n = 7), airway disease (n = 15).
50 patients were further evaluated by
echocardiogram and pulmonary function test
Another 11 cases were excluded due to left ventricular
ejection fraction < 40% (n = 6), severe mitral valve
regurgitation (n = 4) and severe aortic regurgitation
(n = 1)
Gr.A (Ppa < 35 mmHg), n = 15 Gr.B (Ppa ≥ 35 mmHg), n = 24
39 eligible patients were enrolled in the study
Fig. 1. The flow chart of the patient selection process.
Methods
Study design
A total of 97 uraemic patients undergoing maintenance haemodialysis in
our hospital were evaluated for eligibility. This study was approved by the
institution review board and each participant signed an informed consent
form before entering the study. Patients with comorbid conditions that
could cause secondary PH were excluded. The exclusion criteria included
left heart diseases (left ventricle systolic ejection fraction <40%, valvular
heart diseases, coronary artery disease); lung diseases that lead to sec-
ondary PH (asthma, chronic obstructive pulmonary diseases, pulmonary
embolism, etc.); collagen vascular diseases (scleroderma, systemic lupus
erythematous, mixed connective tissue disease); a current smoker or a
smoking history of more than five packs of cigarettes per year, corticos-
teroid therapy, history of respiratory infection in the previous 6 weeks,
pregnancy, human immunodef iciency virus (HIV) infection or the use of
appetite suppressant medication. The patient selection process is depicted
in Figure 1. After an initial review of the medical charts, 47 patients were
excluded due to the past history of coronary heart disease (n=25), au-
toimmune diseases (n=7) and airway disease (n=15). The remaining
50 patients were further subjected to a second-step evaluation by cardi-
ologist and chest medicine experts. A thorough pulmonary function test
and an echocardiogram were performed for each patient. Echocardiogra-
phy was performed for each patient to assess systolic and diastolic left
ventricle function and pulmonary artery pressure by a single physician.
Ejection fractions were used as an index of systolic function. Patients
with Ppa >35 mmHg were further evaluated by two experienced chest
specialists in order to uncover other potential causes of PH. This assess-
ment included history, physical examination, chest radiography, complete
pulmonary function tests and measurements of arterial blood gases. Pul-
monary function was evaluated with body plethysmography. All tests were
performed according to the American thorax society standards with pa-
tients in a sitting position at the same time of day and performed by the
same technician in order to ensure the consistency of a technique. Eleven
patients were excluded due to left ventricular ejection fraction <40%
(n=6), severe mitral valve regurgitation (n=4) and severe aortic valve
regurgitation (n=1). The remaining 39 eligible cases were then divided
into group A (n=15), who did not have PH (Ppa<35 mmHg), and group
B(n=24), whose Ppa ≥35 mmHg.
Assessment of pulmonary artery pressure
Systolic Ppa was assessed by Doppler echocardiography using an ul-
trasound machine equipped with 2.5 MHz probe (Philips Sonos 5500,
Andover, MA, USA) by a single investigator blinded to the results of
the biochemical analyses. To avoid overestimation of Ppa due to volume
overload, the echocardiology studies in patients receiving haemodialysis
were performed within 1 h of completion of dialysis while patients were at
an optimal dry weight. Complete two-dimensional, M-mode and Doppler
echocardiographic studies were obtained from each patient. A tricuspid
regurgitation systolic jet was recorded from the parasternal or apical win-
dow using a continuous-wave Doppler probe. Systolic right ventricular (or
pulmonary artery) pressure was calculated using the modified Bernoulli
equation: Ppa =4×(tricuspid systolic jet)2+10 mmHg (estimated right
artrial pressure). The accuracy of systolic Ppa estimation by Doppler in
the laboratory has been shown to be accurate and to correlate well (up to
97%) with haemodynamic measurements during right heart catheterization
in patients [12]. The study from the consensus symposium on PH held in
Venice, Italy, in 2003 suggested that a systolic Ppa of 35 mmHg represents
the cutoff value for PH when assessed by Doppler echocardiography [13].
Measurement of FENO
FENO was measured using a chemiluminescence analyser (Sievers Model
280 NOA, Sievers Instrument Inc., Boulder, CO, USA) according to ATS
recommendation [14]. This was performed for the 39 patients by an ex-
perienced technician who was blind to the results of Ppa. Before the
examination of FENO, patients were subjected to the restriction of intake
of food containing nitrite for 3 days, discontinued use of medications such
as nitrite drugs, angiotensin-converting enzyme inhibitors, angiotensin II
receptor antagonists and calcium channel blockers to eliminate possible
confounding factors. All patients fasted for at least 12 h before examina-
tion. The nitric oxide (NO) concentration in exhaled air was determined at
rest in a sitting position over 3 h. In patients who underwent haemodialysis,
the NO concentration in exhaled air was measured 30 min before and after
haemodialysis. During the haemodialysis session, heparin was not used in
order to avoid any interference with the measurement of FENO.
The procedure was performed as previously described [15]. The instru-
ment was not switched off and was calibrated each morning. The calibrated
gas contained carbon oxide (<18 ppm) and nitrogen (<15 ppm). When
measuring FENO, the subject was asked to perform a slow vital capacity
manoeuvre for 30 s against a fixed respiratory resistance. The pressure
level during exhalation was optimized by following the computer screen
on-line to reach a constant 50 ml s−1flow rate. Exhaled air was led through
a non-rebreathing valve into a Teflon tubing system connected to the anal-
yser. The standard deviation between three exhaled samples was <5%
and the detection limit was 1 part per billion (ppb). The subjects inhaled
normal room temperature air to total lung capacity (TLC) via their mouths
and exhaled nasally while targeting a flow signal displayed on a computer
monitor. Expiration continued until a steady NO levellasting for at least 10
s was reached. The contribution of oral NO was excluded. The measure-
ment was repeated three times and then the mean value was calculated. In
the examination of FENO, 69 healthy volunteers matching in age, gender
and body mass index were selected and served as a normal control group.
Measurement of plasma high sensitivity-C reactive protein (hs-CRP) and
cytokines
Plasma high sensitivity-Creactive protein (hs-CRP) was determined
by means of the particle-enhanced immunonephelometry method (Dade
Behring, Germany). Serum cytokines (IL-1β,2,4,5,6,8,10,12,TNF-α,
TNF-βand INF-γ) were measured using Th1/Th2 11 plex FlowCytomix
Multiplex kits (Bender MedSystems, BMS810FF human, Vienna, Aus-
tria) by the cytometric beads assay. The patients’ sera for the study of
cytokines were obtained before sessions of haemodialysis and measured
immediately in the laboratory in our hospital. The patients’ general data
(age, sex, co-morbidities, medication used) and data regarding kidney dis-
ease (aetiology of renal failure, duration of haemodialysis therapy and
vascular access location) were obtained from medical charts. The dialy-
sis regimen including high-flux membranes (Fresenius FX80, Fresenius
Medical Care, Bad Homburg, Germany) and bicarbonate-based dialysis
fluid (Na+: 140.0 mEq/l, K+: 2.0 mEq/l, Ca2+: 2.5, 3.0, 3.5 mEq/l, Mg2+:
1.0 mEq/l, Cl−: 107.0 Eq/l,CH3COO−: 5.0 mEq/l, HCO3−: 39.0 mEq/l,
dextrose: 100 mg/dl) was equally prescribed in both groups. During the
sessions of dialysis, dialysate flowrates ranged from 500 to 700 ml/min and
blood flow rates ranged from 200 to 300 ml/min and adjusted depending
on the clinical condition. Water microbiological testing was determined
by using the tryptone glucose extract agar medium within the incubation
temperature of 17–23◦C for 7 days. The endotoxin level in dialysis wa-
ter was measured by using PYROTELL R
, Gel-clot formulation, G5250,
Cape Cod, USA, whose sensitivity range was 0.25 EU/ml. Biochemical
data including haematocrit, serum calcium, phosphate, intact-parathyroid
by guest on November 3, 2015http://ndt.oxfordjournals.org/Downloaded from
1948 T. M. Yu et al.
Tab l e 1 . Demographic characteristics in group A and group B
Group A Group B
Clinical characteristics (n=15) (n=24) P-value
Ppa (mmHg) 28.8 ±1.2 42.7 ±0.9 <0.001
Gender (male/female) 6/9 10/14 NS
Age (years) 61.0 ±3.8 51.2 ±2.8 NS
Aetiology of renal failure NS
Chronic glomerulonephritis 7 13
Hypertension 4 3
Diabetic mellitus 4 8
Medications of anti- NS
hypertension
Calcium channel blocker 11 17
ACEI /ARB 6 8
Beta blockers 9 13
Alpha blockers 4 5
Duration of dialysis (ms) 75.3 ±12.9 72.5 ±13.1 NS
Microorganisms (CFU/ml) 1.7 ±0.9 3.5 ±1.6 NS
Endotoxin (>0.25 IU/ml) 0 2 NS
BMI∗(kg/m2) 25.8 ±1.1 24.2 ±0.7 NS
LV ejection fraction (%) 55.7 ±1.7 57.6 ±1.1 NS
Cardiac output (l/min) 5.6 ±0.2 6.8 ±0.3 NS
Access flow (ml/min) 832.8 ±66.1 1350.4 ±74.3 <0.001
Brachial/radial 1.0/14.0 2.0/22.0
Cardiothoracic ratio 0.48 ±0.04 0.56 ±0.08 <0.0001
BMI =body mass index; NS =insignificant.
hormone, arterial blood gases, lipid profile, albumin, uric acid and ferritin
were also retrieved. The mean of the 12 monthly values preceding the
echocardiography study was utilized.
Measurement of vascular access blood flow
Arterio-venous fistula blood flow was measured using a Transonic
Haemodialysis Monitor (Transonic System Inc., Ithaca, NY, USA). All
measurements were performed in triplicate at a fixed dialyzer blood flow
rate of 250 ml/min within the first hour after the start of haemodialysis.
Ultrafiltration was turned off during blood flow measurements. The mean
value of three results was used.
Statistical analyses
Continuous variables were expressed as mean ±SD unless otherwise
specified. Variables of cytokines which were not normally distributed
were expressed as mean ±standard error of mean (SEM).
The subgroups were compared using the independent t-test and Mann–
Whitney U-test. Pearson correlation analysis was used to explain the re-
lationship between two parameters. Categorical variables, expressed as
percentages, were analysed by Fisher’s exact test or Yate’s correction of
contingency. All statistical tests were two-sided and assessed at a 0.05
significance level. Analyses were performed using the Scientific Package
for Social Science (SPSS, version 10.1, Chicago, IL, USA).
Results
PH was found in 24 (61.53%) cases (group B). The mean
Ppa was 42.7 ±4.6 mmHg in this group. The other group
(15 cases, group A) had no PH and the mean Ppa was 28.8 ±
4.8 mmHg (P<0.0001, compared to group B). Demo-
graphic characteristics and biochemical data are listed and
compared in Tables 1 and 2. Group B was younger (51.2 ±
13.8 versus 61.0 ±14.7, P=0.041) and had lower
solute clearance and residual renal function (Kt/V: 1.4 ±
0.2 versus 1.7 ±0.2, P<0.001, urea reduction rate: 69.3 ±
4.5 versus 75.0 ±2.4, P<0.001 and residual Ccr ml/min:
2.0 ±0.1 versus 4.2 ±0.1, P<0.0001). In addition, a
Tab l e 2 . Biochemical variables in group A and group B
Group A Group B
Variables (n=15) (n=24) P-value
Albumin (g/dl) 3.97 ±0.08 3.68 ±0.08 0.038
HbA1c (%) 5.7 ±0.2 7.0 ±0.2 <0.001
Ferritin (ng/ml) 434.9 ±128.2 279.1 ±55.8 NS
Hct (%) 32.1 ±0.7 31.7 ±0.9 NS
Ca ×phosphate 46.3 ±3.0 52.0 ±2.6 NS
I-PTH (pg/ml) 267.9 ±99.8 307.6 ±58.1 NS
Cholesterol (mg/dl) 170.6 ±8.2 179.4 ±5.8 NS
Triglyceride (mg/dl) 169.9 ±22.3 165.9 ±16.6 NS
Uric acid (mg/dl) 7.2 ±0.2 6.8 ±0.3 NS
HCO3−(mmol/l) 24.0 ±0.9 23.0 ±0.6 NS
Residual Ccr (ml/min) 4.2 ±0.1 2.0 ±0.1 <0.0001
Kt/V 1.7 ±0.0 1.4 ±0.0 <0.001
URR 75.0 ±0.6 69.3 ±0.9 <0.001
NS =insignificant.
Ppa (mmHg)
access blood flow (ml/min)
15 20 25 30 35 40 45 50 55
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
male female
n = 19, r = 0.553, P = 0.014
n = 17, r = 0.861, P < 0.0001
Fig. 2. Moderate correlation (r=0.61) was noted between access blood
flow and Ppa in total cases. Compared to the male population, higher
correlation (r=0.861) was found in the female population.
significantly higher cardiac output was found in group B
compared to group A (6.8 ±1.7 l/min versus 5.6 ±0.7
l/min, P=0.014). Vascular access blood flow was also sig-
nificantly higher in group B compared to g roup A (1350.4 ±
364.1 versus 832.8 ±255.8 ml/min, P<0.001). The im-
pact of vascular access flow on Ppa in different genders
and the relationship between access blood flow and Ppa are
shown in Figure 2. Patients with higher blood flow showed
increased Ppa irrespective of the gender. The serum levels
of hs-CRP and various kinds of cytokines including IL-1,
2, 4, 5, 6, 8, 10, 12, TNF-α,TNF-βand INF-γall revealed
a higher level in group B than those in group A (Table 3).
In group B (Figure 3), the serum level of hs-CRP (4.70 ±
0.92 versus 1.48 ±0.41 mg/l, P=0.001) and inflammatory
cytokines including IL-1β(2.24 ±0.10 versus 1.65 ±0.04
pg/ml, P<0.0001), TNF-α(5.51 ±1.55 versus 2.60 ±0.11
pg/ml, P=0.028) and IL-6 (14.54 ±0.10 versus 14.16 ±
0.10 pg/ml, P=0.011) reached a statistical significance
when compared to group A.
FENO levels are shown in Figure 4. Pre-dialysis FENO
levels were significantly higher in both uraemia group B
by guest on November 3, 2015http://ndt.oxfordjournals.org/Downloaded from
Systemic inflammation is associated with PH 1949
0
5
10
15
20
25
AB
hs-CRP (mg/L)
P = 0.001
13.5
14.0
14.5
15.0
15.5
16.0
16.5
P = 0.011
AB
IL-6 (pg/mL)
13.0
IL-1β (pg/mL)
1.0
1.5
2.0
2.5
3.0
3.5
4.0
AB
P < 0.0001
TNF-α (pg/mL)
0
10
20
30
40
50 P = 0.028
AB
Fig. 3. Cytokines and hs-CRP were significantly different in groups A and B.
Tab l e 3 . Plasma levels of hs-CRP and cytokines
Group A (n=15) Group B (n=24) P-value
hs-CRP (mg/l) 1.48 ±0.41 4.70 ±0.92 0.001
IL-6 (pg/ml) 14.16 ±0.10 14.54 ±0.10 0 011
IL-1β(pg/ml) 1.65 ±0.04 2.24 ±0.10 <0.0001
TNF-α(pg/ml) 2.60 ±0.11 5.51 ±1.55 0.028
IL-2 (pg/ml) 1.84 ±0.03 2.09 ±0.14 0.248
IL-4 (pg/ml) 1.34 ±0.03 1.46 ±0.08 0.896
IL-5 (pg/ml) 1.64 ±0.02 2.58 ±0.97 0.696
IL-8 (pg/ml) 16.01 ±3.35 16.13 ±1.54 0.453
IL-10 (pg/ml) 5.43 ±0.12 5.45 ±0.11 0.851
IL-12 (pg/ml) 1.19 ±0.02 1.29 ±0.07 0.593
IFN-γ(pg/ml) 1.92 ±0.06 1.95 ±0.06 1.000
Data were expressed as mean ±SEM (standard error of mean).
(39.9 ±16.7 ppb, P<0.0001) and group A (31.8 ±
10.3 ppb, P=0.001) compared to healthy control lev-
els (19.8 ±6.8 ppb), respectively. A further comparison
between the two uraemic groups revealed that the pre-
dialysis FENO levels in group B were significantly higher
than those in group A (39.9 ±16.7 versus 31.8 ±10.3
ppb, P=0.045). A significantly positive correlation was
noted between FENO and these four plasma cytokines in-
cluding hs-CRP (r=0.864, P<0.0001), IL-6 (r=0.746,
P<0.0001), IL-1 (r=0.801, P<0.0001) and TNF-α(r=
0.653, P<0.0001). Post-dialysis FENO levels remained sig-
nificantly higher in group B compared to patients in group
A (30.3 ±10.4 ppb versus 20.1 ±10.9 ppb, P=0.003) and
normal controls (P<0.0001). On the other hand, the post-
dialysis level of FENO in group A returned to a normal range
and was not significantly different from that of normal
controls.
Discussion
PH is an atherosclerotic disease that may occur with un-
known aetiology (primary PH) or secondary to another
disease (secondary PH). The pathological findings of in-
flammatory cells’ infiltration around plexiform lesions and
the association of PH with several chronic inflammatory
diseases [16–19] suggest that immunological mechanisms
may be involved in the pathogenesis of PH [2,7,18]. In
many in vitro studies, interleukins, such as IL-1β,IL-6,
and TNF-α, have been identified and elevated in local lung
tissue surrounding pulmonary arteries in patients with PH
[2,18–19]. Uraemia has been regarded as a state of chronic
inflammation, in which elevated serum levels of several cy-
tokines and growth factors have been reported previously
[21–22]. Many studies have reported increased production
of various cytokines including IL-1β,TNF-α,IL-1β,IL-
6, IL-8, etc. in HD, CAPD and undialyzed patients with
end-stage renal failure and lead to a complex of acute and
chronic side effects [23]. Hence, it is not surprising that PH
by guest on November 3, 2015http://ndt.oxfordjournals.org/Downloaded from
1950 T. M. Yu et al.
0
10
20
30
40
50
FENOPPb
Pre- Post-
Group A
Pre- Post-
Group B
Normal
P < 0.0001
P < 0.0001
P = 0.001 P = 0.045
P = 0.003
19.9
31.8
20.1
39.9
30.3
Fig. 4. The pre- and post-dialysis levels of FENO in group A, group
B and healthy controls. Higher pre-dialysis FENO levels were found in
patients undergoing haemodialysis. Compared with group A, the pre- and
post-dialysis levels of group B were significantly higher.
is relatively common among chronically dialyzed patients
with a prevalence that ranged from 40 to 50% in previous
reports [9–11]. In patients with PH in our study, poorer
underlying conditions including lower residual renal func-
tion, poorer clearance rate and higher HbA1c indirectly
reflect the severity of inflammation in these patients. In
an attempt to clarify whether PH in uraemia occurs as a
consequence of systemic inflammation, we carried out two
parts of the study. The first part measured several biomark-
ers, including hs-CRP, IL-1β, 2, 4, 5, 6, 8, 10, 12, TNF-
α,TNF-β,INF-γ, and clearly demonstrated that group B
(with PH) patients had significantly higher serum inflam-
matory biomarker levels of hs-CRP, IL-1β,TNF-αand IL-6
(Figure 3). The second part measured FENO in both groups
and controls. FENO has been extensively used in studies
of various airway inflammatory diseases and is considered
as a biomarker of inflammation in lung tissue [14,20]. In
the present study, we have demonstrated that uraemic pa-
tients, no matter if PH existed, had significantly higher
pre-dialysis FENO (Figure 4), implying that local inflam-
mation in the lung tissue prevails in dialyzed patients. The
inflammation is more severe in group B patients as reflected
by significantly higher FENO levels compared to group A
patients. Furthermore, a significantly positive correlation
existed between plasma cytokines and airway FENO ,im-
plying that systemic inflammation as well as local airway
inflammation contributes to the pathogenesis of PH consis-
tently. After dialysis, FENO declined to a level comparable
with normal controls in group A patients. The finding that
group B patients had significantly higher levels of FENO
even after dialysis suggests a persistently intensive inflam-
mation in the airway of these patients with PH, which is
compatible with the assumption that inflammation plays
an important role in the pathogenesis of PH in uraemic
patients.
This is further supported by the findings of elevated
blood levels of various inflammatory biomarkers in this
group of patients. The interacting mechanism of increased
endogenous NO through up-regulation of i-NOS regu-
lated by inflammatory cytokine such as IL-1 and TNF in
chronic haemodialysis patients has been clearly demon-
strated and elucidated in the literature [24–26]. Amore
et al. [25] have shown that abnormal stimulation of i-
NOS through cytokines was closely associated with the
development of long-term vasculopathy in dialysis pa-
tients that is consistent with the assumption in our study.
Whether peroxynitrite, a strong oxidizing agent produced
from NO and superoxide anion involved with the patho-
genesis of cardiovascular disorders in uraemia, could con-
tribute to the pathogenesis of PH in uraemia in our result
is subjected to further investigation in the future [27]. In
addition, we demonstrated a trend towards an increased
prevalence of pulmonary hypertension (PAH) when the vas-
cular access flow increased. The trend reached the statisti-
cal significance and was more obvious in female patients
(Figure 2). The different influence of access blood flow on
PAH in male and female populations in our results was ob-
vious but not readily explained and requires further explo-
ration. We suggest that a special attention must be paid, not
only to the effects of access flow on the left heart but also to
the impact on pulmonary artery pressure in haemodialysis
patients. The clinical significance of our work is clear: first,
a successful renal transplantation is likely to ameliorate PH
through the recovery of uraemia, which is inflammatory
per se. Indeed, renal transplantation has been found to
markedly improve the stiffness of the vascular structure
and to normalize PH [28,29]. Secondly, intensive attempts
are needed to reduce systemic inflammation of dialysis pa-
tients. Although dialysis water microbiological data are not
significantly different in both groups, we consider that using
a more biocompatible dialyzer and/or ultrapure dialysate
may help in reducing the degree of PH, but this warrants
more studies for confirmation.
In conclusion, the pathogenesis of PH in haemodialysis
is complex and the results of the study raise the possibility
of a novel pathogenetic role of systemic inflammation in
the pathogenesis of PH. Due to the limitation of small size
and cross-section in our study, larger longitudinal studies
are needed to further address the question in the future.
Acknowledgements. We thank the biostatistics task force of Taichung
Veterans General Hospital for assistance in statistical analysis and Dr Bor-
Jen Lee for excellent support of cardiac echocardiogram. We also thank
Dr Chen-Yu Wong for the assistance to assess pulmonary morbidity in the
study.
Conflict of interest statement. The study received a research grant from
Taichung Veterans General Hospital, Taiwan (TCVGH-973601A).
References
1. Archer S, Rich S. Primary pulmonary hypertension: a vascular biology
and translational research “Work in progress.” Circulation 2000; 102:
2781–2791
2. Dorfmuller P, Perros F, Balabanian K et al. Inflammation in pul-
monary arterial hypertension. EurRespirJ2003; 22: 358–363
by guest on November 3, 2015http://ndt.oxfordjournals.org/Downloaded from
Systemic inflammation is associated with PH 1951
3. Tuder RM, Groves B, Badesch DB et al. Exuberant endothelial cell
growth and elements of inflammation are present in plexiform lesions
of pulmonary hypertension. Am J Pathol 1994; 144: 275–285
4. Cool CD, Kennedy D, Voelkel NF et al. Pathogenesis and evolution
of plexiform lesions in pulmonary hypertension associated with scle-
roderma and human immunodeficiency virus infection. Hum Pathol
1997; 28: 434–442
5. Kimura H, Kasahara Y, Kurosu K, et al. Alleviation of monocrotaline-
induced pulmonary hypertension by antibodies to monocyte chemo-
tactic and activating factor/monocyte chemoattractant protein-1. Lab
Invest 1998; 78: 571–581
6. Fujita M, Shannon JM, Irvin CG et al. Overexpression of tumor necro-
sis factor-alpha produces an increase in lung volumes and pulmonary
hypertension. Am J Physiol Lung Cell Mol Physiol 2001; 280: L39–
L49
7. Humbert M, Monti G, Brenot F et al. Increased interleukin-1 and
interleukin-6 serum concentrations in severe primary pulmonary hy-
pertension. AmJRespirCritCareMed1995; 151: 1628–1631
8. YiglaM, Keidar Z, Safadi I et al. Pulmonary calcif ication in hemodial-
ysis patients: correlation with pulmonary artery pressure values. Kid-
ney Int 2004; 66: 806–810
9. Abassi Z, Nakhoul F, Khankin E et al. Pulmonary hypertension in
chronic dialysis patients with arteriovenous fistula: pathogenesis and
therapeutic prospective. Curr Opin Nephrol Hypertens 2006; 15: 353–
360
10. Nakhoul F, Yigla M, Gilman R et al. The pathogenesis of pulmonary
hypertension in haemodialysis patients via arterio-venous access.
Nephrol Dial Transplant 2005; 20: 1686–1692
11. Havlucu Y, Kursat S, Ekmekci C et al. Pulmonary hypertension in
patients with chronic renal failure. Respiration 2007; 74: 503–510
12. Eysmann SB, Palevsky HI, Reichek N et al. Two-dimensional and
Doppler-echocardiographic and cardiac catheterization correlates of
survival in primary pulmonary hypertension. Circulation 1989; 80:
353–360
13. Barst RJ, McGoon M, Torbicki A et al. Diagnosis and differential
assessment of pulmonary arterial hypertension. J Am Coll Cardiol
2004; 43: S40–SS47
14. ATS/ERS recommendations for standardized procedures for the on-
line and offline measurement of exhaled lower respiratory nitric oxide
and nasal nitric oxide, 2005. Am J Respir Crit Care Med 2005; 171:
912–930
15. Depalo A, Carpagnano GE, Spanevello A et al. Exhaled NO and
iNOS expression in sputum cells of healthy, obese and OSA subjects.
J Intern Med 2008; 263:70–78
16. Lesprit P, Godeau B, Authier FJ et al. Pulmonary hypertension in
POEMS syndrome: a new feature mediated by cytokines. Am J Respir
Crit Care Med 1998; 157: 907–911
17. Nishimaki T, Aotsuka S, Kondo H et al. Immunological analysis of
pulmonary hypertension in connective tissue diseases. J Rheumatol
1999; 26: 2357–2362
18. Dorfmuller P, Zarka V, Durand-Gasselin I et al. Chemokine RANTES
in severe pulmonary arterial hypertension. AmJRespirCritCareMed
2002; 165: 534–539
19. Joppa P, Petrasova D, Stancak B et al. Systemic inflammation in
patients with COPD and pulmonary hypertension. Chest 2006; 130:
326–333
20. Kaneko FT, Arroliga AC, Dweik RA et al. Biochemical reaction
products of nitric oxide as quantitative markers of primary pul-
monary hypertension. Am J Respir Crit Care Med 1998; 158: 917–
923
21. Pertosa G, Grandaliano G, Gesualdo L et al. Clinical relevance of cy-
tokine production in hemodialysis. Kidney Int Suppl 2000; 76: S104–
S111
22. Pereira BJ, Shapiro L, King AJ et al. Plasma levels of IL-1 beta,
TNF alpha and their specific inhibitors in undialyzed chronic renal
failure, CAPD and hemodialysis patients. Kidney Int 1994; 45: 890–
896
23. Henderson LW, Koch KM, Dinarello CA et al. Haemodialy-
sis hypotension: the interleukin hypothesis. Blood Purif 1983; 1:
3–8
24. Madore F, Prud’homme L, Austin JS et al. Impact of nitric oxide on
blood pressure in hemodialysis patients. Am J Kidney Dis 1997; 30:
665–671
25. Amore A, Bonaudo R, Ghigo D et al. Enhanced production of nitric
oxide by blood-dialysis membrane interaction. J Am Soc Nephrol
1995; 6: 1278–1283
26. Lamas S, Michel T, Brenner BM et al. Nitric oxide synthesis in
endothelial cells: evidence for a pathway inducible by TNF-alpha. Am
JPhysiol1991; 261: C634–C641
27. Weinberger B, Fakhrzadeh L, Heck DE et al. Inhaled nitric ox-
ide primes lung macrophages to produce reactive oxygen and ni-
trogen intermediates. AmJRespirCritCareMed1998; 158: 931–
938
28. Covic A, Goldsmith DJ, Gusbeth-Tatomir P et al. Successful renal
transplantation decreases aortic stiffness and increases vascular reac-
tivity in dialysis patients. Transplantation 2003; 76: 1573–1577
29. Yigla M, Nakhoul F, Sabag A et al. Pulmonary hypertension in pa-
tients with end-stage renal disease. Chest 2003; 123: 1577–1582
Received for publication: 12.6.08; Accepted in revised form: 15.12.08
by guest on November 3, 2015http://ndt.oxfordjournals.org/Downloaded from
