Pulmonary hypertension in peritoneal dialysis patients.

Abstract

The information available in the literature regarding pulmonary hypertension (PH) in peritoneal dialysis (PD) patients is limited. The objective of the present study was to examine the prevalence and characteristics of PH in PD patients. We retrospectively collected the clinical profile, echocardiographic (ECHO) findings, and biochemical data for 36 PD patients for which ECHO findings were available. We compared characteristics between patients with and without PH. We found PH, defined as pulmonary arterial pressure (PAP) > or = 35 mmHg, in 15 patients. The prevalence of PH was 42%. Mean age (+/- standard deviation) of the patients with and without PH was 58 +/- 15 years and 52 +/- 15 years respectively (p = 0.30). Mean PAP of the PH patients was 43.8 +/- 9.0 mmHg (range: 35-65 mmHg). Patients with PH had a lower ejection fraction than did patients without PH (46.3% +/- 19.8% vs. 56.5% +/- 11.8% respectively, p = 0.07). Patients with PH also had a higher prevalence of global hypokinesia (60% vs. 29%, p = 0. 059) and dilated left ventricular chamber (53% vs. 19%, p = 0.03). In PH patients, body mass index (24 +/- 4.5 kg/m2 vs. 28 +/- 5.0 kg/m2, p = 0.024), normalized protein catabolic rate (0. 78 +/- 0.21 g/kg vs. 0.95 +/- 0.27 g/kg daily, p = 0.049), and ferritin (226 +/- 210 ng/mL vs. 873 +/- 965 ng/mL, p = 0.005) were significantly lower and lactate dehydrogenase was higher (264 +/- 99 U/L vs. 206 +/- 79 U/L, p = 0.06) than in patients without PH. We observed no significant differences in race or sex, incidence of hypertension or cardiovascular disease, or vitamin D analog use between the two groups of patients. During the study period, 60% of PH patients and 38% of patients without PH died (p = 0.19). Values of PAP correlated directly with serum levels of phosphorus (r = 0.44, p = 0.02), CaxP product (r = 0.40, p = 0.04), and parathyroid hormone (r = 0.42, p = 0.03). Of continuous ambulatory PD and continuous cycling PD patients, 21% and 55% respectively had PH (p = 0. 049). In PD patients, PH is highly prevalent and may be associated with higher mortality risk.

Full text

Page 1

Advances in Peritoneal Dialysis, Vol. 23, 2007
The information available in the literature regarding
pulmonary hypertension (PH) in peritoneal dialysis
(PD) patients is limited. The objective of the present
study was to examine the prevalence and character-
istics of PH in PD patients.
We retrospectively collected the clinical profile,
echocardiographic (ECHO) findings, and biochemi-
cal data for 36 PD patients for which ECHO findings
were available. We compared characteristics between
patients with and without PH. We found PH, defined
as pulmonary arterial pressure (PAP) ≥ 35 mmHg, in
15 patients. The prevalence of PH was 42%. Mean
age (± standard deviation) of the patients with and
without PH was 58 ± 15 years and 52 ± 15 years re-
spectively (p = 0.30). Mean PAP of the PH patients
was 43.8 ± 9.0 mmHg (range: 35 – 65 mmHg).
Patients with PH had a lower ejection fraction
than did patients without PH (46.3% ± 19.8% vs.
56.5% ± 11.8% respectively, p = 0.07). Patients with
PH also had a higher prevalence of global hypoki-
nesia (60% vs. 29%, p = 0.059) and dilated left ven-
tricular chamber (53% vs. 19%, p = 0.03). In PH
patients, body mass index (24 ± 4.5 kg/m2 vs. 28 ±
5.0 kg/m2, p = 0.024), normalized protein catabolic
rate (0.78 ± 0.21 g/kg vs. 0.95±0.27 g/kg daily, p =
0.049), and ferritin (226 ± 210 ng/mL vs. 873 ±
965 ng/mL, p = 0.005) were significantly lower and
lactate dehydrogenase was higher (264 ± 99 U/L vs.
206 ± 79 U/L, p = 0.06) than in patients without PH.
We observed no significant differences in race or sex,
incidence of hypertension or cardiovascular disease,
or vitamin D analog use between the two groups of
patients. During the study period, 60% of PH patients
and 38% of patients without PH died (p = 0.19).
Values of PAP correlated directly with serum levels
of phosphorus (r = 0.44, p = 0.02), Ca×P product
(r = 0.40, p = 0.04), and parathyroid hormone (r =
0.42, p = 0.03). Of continuous ambulatory PD and
continuous cycling PD patients, 21% and 55% re-
spectively had PH (p = 0.049). In PD patients, PH
Pulmonary Hypertension in
Peritoneal Dialysis Patients
Lalathaksha Kumbar, Paul A. Fein, Muhammad A. Rafiq,
Cezary Borawski, Jyotiprakas Chattopadhyay, Morrell M.
Avram
is highly prevalent and may be associated with higher
mortality risk.
Key words
Pulmonary hypertension, parathyroid hormone, PTH,
pulmonary arterial pressure, mortality
Introduction
Pulmonary hypertension (PH) is increasingly being
recognized as a factor that can affect outcome in end-
stage renal disease (ESRD). Most cases of PH are
consequent to myocardial or pulmonary diseases.
Chronic left ventricular failure and the resulting el-
evation in pulmonary venous pressure is the most
common cause of PH. In several populations without
chronic kidney disease (CKD), PH has been associ-
ated with excess mortality (1–3). A high prevalence
of PH in CKD patients on hemodialysis (HD) has re-
cently been reported (4–7). In a retrospective study
(8), we reported PH in 64% of HD patients who un-
derwent echocardiography (ECHO) for any reason
within the preceding 12 months (36% of our entire
HD population, reflecting the minimum prevalence).
End-stage renal disease commonly occurs with
significant comorbidities, several of which may lead
to PH. Left-side heart disease, chronic obstructive
pulmonary disease, sleep apnea, collagen vascular
diseases, HIV infection, and portal hypertension may
coexist and result in PH. Most HD patients have an
arteriovenous (AV) hemoaccess that results in AV
shunting and potentially in PH. In addition, second-
ary hyperparathyroidism, elevated Ca×P product, and
vascular calcification are commonly seen in CKD
patients (9,10).
In an animal model, Akmal et al. showed that ex-
cess parathyroid hormone (PTH) in chronic renal fail-
ure induces pulmonary calcification, PH, and right
ventricular hypertrophy (11). Recently, we reported
the prevalence and clinical characteristics of PH in
CKD patients before initiation of renal replacement
therapy (12).
The information available in the literature regard-
ing PH in patients treated with peritoneal dialysis (PD)
From: Avram Division of Nephrology, Long Island College
Hospital, Brooklyn, New York, U.S.A.

Page 2

128 Kumbar et al.
is limited (4). The objective of the present study was
therefore to examine the prevalence, clinical profile,
and echocardiography characteristics of PH in PD
patients and to identify the biochemical markers as-
sociated with PH in PD patients.
Patients and methods
We retrospectively reviewed the medical records of
80 PD patients treated at the Avram Center for Kid-
ney Diseases (the outpatient facility at Long Island
College Hospital) between 2002 and 2005. Of these
patients, 36 had undergone ECHO.
On patient enrollment into the study, we collected
demographic data (including age, race, and sex) and
clinical data (including body weight, body mass in-
dex, presence of diabetes and hypertension, known
cardiovascular or pulmonary disease, medication his-
tory, and tobacco and alcohol use). Biochemical
parameters and ECHO findings were collected for
each patient. Laboratory data nearest to the patient’s
ECHO study were included in the analysis. We de-
fined PH as a systolic pulmonary artery pressure
(PAP) ≥ 35 mmHg (4). For 8 patients whose PAP
was not recorded, we presumed a lack of PH.
We compared characteristics between patients
with and without PH. The Institutional Review
Board of Long Island College Hospital approved
the study protocol.
Statistical analyses
Continuous variables are reported as mean ± standard
deviation. For selected comparisons between two
group means, the parametric (t-test) or nonparamet-
ric (Mann–Whitney test) was used, as applicable.
Categorical variables were compared using the chi-
square test or the Fisher exact test, as appropriate.
Correlations were reported either as Pearson correla-
tion coefficients or Spearman rank correlation coeffi-
cients. Calculations were performed using SPSS
12.0.1 for Windows (SPSS Inc., Chicago, IL. U.S.A.).
Results
Of the 36 PD patients, 15 had PH, a prevalence of
42%. For the groups without and with PH, we ob-
served no differences in age (52 ± 15 years vs. 58 ±
15 years, p = 0.30), sex (women: 62% vs. 53%), race
(African American: 71% vs. 73%), or cause of CKD.
Table I compares the clinical characteristics of
patients without and with PH. Surprisingly, the patients
without PH had a higher prevalence of diabetes (57%
vs. 20%, p = 0.04) and peripheral vascular disease
(29% vs. 0%, p = 0.03). Patients with PH had higher
prevalences of congestive heart failure (CHF: 13%
vs. 0%; p = 0.08) and coronary artery disease (CAD:
40% vs. 14%; p = 0.08). The percentage of smokers
was higher among PH patients than among those with-
out PH (53% vs. 24%, p = 0.07). During the study
period, 60% of the PH patients and 38% of the pa-
tients without PH died (p = 0.31), highlighting the
risk accompanying PH is this population.
Table II shows biochemical parameters for pa-
tients without and with PH. Patients with PH had a
significantly lower normalized protein catabolic rate
(0.78 g/kg vs. 0.95 g/kg, p = 0.049), significantly
lower ferritin (227 ng/mL vs. 873 ng/mL, p = 0.008),
and higher lactic dehydrogenase (LDH: 264 U/L vs.
206 U/L; p = 0.06).
Values of PAP were not normally distributed in
PD patients. Mean and median PAP were 37.4 ±
9.96 mmHg and 35 mmHg overall, with a range of
25 – 65 mmHg. Among PH patients, mean PAP was
43.8 ± 9.0 mmHg (range: 35 – 65 mmHg). Values
of PAP correlated positively with serum levels of
phosphorus (r = 0.44, p = 0.02), Ca×P product (r =
0.40, p = 0.04), and parathyroid hormone (PTH:
r = 0.42, p = 0.03; Figure 1), and negatively with
serum ferritin (r = –0.46, p = 0.016).
Table III shows the ECHO characteristics of the
patients without and with PH. Although the p value
did not quite reach significance, patients with PH had
a lower left ventricular ejection fraction than did pa-
tients without PH (46.3% vs. 56.5%, p = 0.07). As
compared with patients without PH, patients with PH
had significantly greater prevalence of dilated left
chamber (53% vs. 19%, p = 0.03) and global hypokin-
esia (60% vs. 29%, p = 0.059).
We also compared the history of medications and
dialysis modality use between the two groups of PD
patients. As expected, because of a higher prevalence
of diabetes, patients without PH were more likely to
be taking insulin (p = 0.01). We observed no signifi-
cant difference between the two groups of PD patients
with regard to treatment with erythropoietin, vitamin
D analogs, phosphate binders, angiotensin convert-
ing-enzyme inhibitors, beta-blockers, and calcium
channel blockers. Of continuous ambulatory PD pa-
tients and continuous cycling PD patients, 21% and
55% respectively had PH (p = 0.049).

Page 3

Pulmonary Hypertension in PD 129
Discussion
The results of the present study indicate a very high
prevalence of PH (42%) on ECHO in our popula-
tion of PD patients having undergone ECHO for
any reason.
To our knowledge, no previous reports regard-
ing the prevalence of PH in PD patients have been
published. In a report concerning 3790 ECHO-nor-
mal subjects, 28% had a systolic PAP greater than
30 mmHg (13). The recently published literature con-
tains a few papers regarding PH in HD patients. Yigla
et al. (4) indicated that 39.7% of their HD patients
had PH (PAP > 35 mmHg). Amin et al. reported a
PH prevalence of 29% in HD patients (5), and most
recently, Tarass reported 27% (7). We recently re-
ported 75% PH prevalence in our pre-dialysis CKD
patients (12), which may in fact reflect pre-existing
disease in PD patients.
In the present study of PD patients, only 1 patient
had a PAP of 65 mmHg—just short of being termed
“severe” PH (PAP > 65 mmHg). In our previously
reported pre-dialysis CKD patients (12), 3 (12.5%)
had severe PH. Our prevalent HD patients have severe
PH prevalence of about 4% [6 of 151 patients by
ECHO (8)]. Yigla et al. reported severe PH in 28% of
the non-ESRD patient population (14).
The lower ferritin levels, higher levels of LDH,
lower normalized protein catabolic rate, and higher
prevalence of CHF and CAD in our PD patients with
PH may contribute to their increased mortality risk.
Reports have indicated that PH is prevalent in adult
patients with sickle cell disease and is strongly asso-
ciated with early mortality and markers of hemolysis,
in particular serum LDH (15). We previously reported
increased mortality risk in HD patients with PH as
compared with patients without PH (8), finding that
TABLE I
Clinical characteristics of peritoneal dialysis patients without and with pulmonary hypertension (PH)
Without PH
(n=21)
With PH
(n=15)
p Value
Body mass index (kg/m2)
Hypertension (%)
Diabetes (%)
Coronary artery disease (%)
Congestive heart failure (%)
HIV-positive (%)
Peripheral vascular disease (%)
Smoking (%)
Residual GFR (mL/min)
Residual renal function (%)
Expiration (%)
28±5.0
90.5
57
14
0
9.5
29
24
0.58±0.86
43
38
24±4.5
93.3
20
40
13
13
0
53
0.70±1.3
27
60
0.024
0.63
0.04
0.08
0.08
0.72
0.03
0.07
0.58
0.32
0.19
GFR = glomerular filtration rate.
TABLE II
Biochemical parameters of peritoneal dialysis patients without and with pulmonary hypertension (PH)
Without PH
(n=21)
With PH
(n=15)
p Value
Hemoglobin (g/dL)
Albumin (g/dL)
Calcium (mg/dL)
Phosphorus (mg/dL)
Parathyroid hormone (pg/mL)
Ferritin (ng/mL)
Transferrin saturation (%)
Lactic dehydrogenase (U/L)
Daily nPCR (g/kg)
10.94±2.48
3.32±0.81
9.86±0.98
4.56±1.8
540±445
873±965
15.9±3.5
206±79
0.95±0.27
10.87±1.91
3.19±0.56
9.79±0.77
5.0±1.8
569±371
226±210
9.74±2.5
264±99
0.78±0.21
0.93
0.58
0.82
0.27
0.84
0.005
0.85
0.06
0.049
nPCR = normalized protein catabolic rate.

Page 4

130 Kumbar et al.
21% of PH patients and 11% of patients without PH
died during the study period. Yigla et al. reported 30%
and 8.5% mortality rates in patients with and without
PH respectively (4). In the present study, however,
the difference in mortality between the two groups of
PD patients did not reach statistical significance. We
could not explain the higher prevalence of PH among
continuous cycling PD patients; that finding needs
further investigation.
Our finding that PTH correlates directly with PAP
in PD patients is interesting. Based on an animal
model, Akmal et al. suggested that excess PTH in
chronic renal failure induces PH, pulmonary calcifi-
cation, and right ventricular hypertrophy, and proposed
that PH develops secondarily to pulmonary artery
calcification. Prevention of pulmonary artery
calcification by parathyroidectomy in dogs with CRF
was associated with normalization of mean PAP and
reduction of right ventricular hypertrophy (11). But
Amin et al. could not find any statistically significant
difference in PTH between 15 patients with PH and
36 patients without PH who were on HD (5).
Conclusions
The present study demonstrates that PH is highly
prevalent in PD patients and may be associated with
higher mortality risk. The PH in our patients prob-
ably reflects a complication of left-chamber heart dis-
ease. The association of hyperparathyroidism with PH
in PD patients is consistent with our findings in preva-
lent HD patients (8). This link needs to be confirmed
and prospectively explored, especially given the
prevalence of secondary hyperparathyroidism, the
large number of PD patients with PH, and the excess
mortality risk seen with PH in this population. Early
diagnosis may enable timely intervention. Consider-
ing the limitations of our small number of patients in
a retrospective study, these findings need to be fur-
ther explored in larger prospective studies.
Acknowledgments
The present work was supported in part by grants from
the Kidney Urology Foundation of America and the
Nephrology Foundation of Brooklyn.
References
1 Abramson SV, Burke JF, Kelly JJ Jr, et al. pulmonary
hypertension predicts mortality and morbidity in pa-
tients with dilated cardiomyopathy. Ann Intern Med
1992;116:888–95.
2 Grigioni F, Potena L, Galie N, et al. Prognostic impli-
cations of serial assessments of pulmonary hyperten-
sion in severe chronic heart failure. J Heart Lung
Transplant 2006;25:1241–6.
TABLE III
Echocardiography characteristics of peritoneal dialysis patients without and with pulmonary hypertension (PH)
Without PH
(n=21)
With PH
(n=15)
p Value
Ejection fraction (%)
Left ventricular hypertrophy (%)
Diastolic dysfunction (%)
Dilated left ventricular chamber (%)
Right ventricular dilatation (%)
Mitral valve calcification (%)
Global hypokinesia (%)
56.5±11.8
43
48
19
4.8
29
29
46.3±19.8
67
33
53
13.3
20
60
0.07
0.16
0.39
0.03
0.56
0.70
0.059
FIGURE 1
roid hormone. Correlation coefficient: r = 0.42; p = 0.03.
Correlation of pulmonary artery pressure with parathy-

Page 5

Pulmonary Hypertension in PD 131
3 Trad S, Amoura Z, Beigelman C, et al. Pulmonary ar-
terial hypertension is a major mortality factor in dif-
fuse systemic sclerosis, independent of interstitial
lung disease. Arthritis Rheum 2006;54:184–91.
4 Yigla M, Nakhoul F, Sabag A, et al. Pulmonary hy-
pertension in patients with end stage renal disease.
Chest 2003;123:1577–82.
5 Amin M, Fawzy A, Hamid MA, Elhendy A. Pulmo-
nary hypertension in patients with chronic renal fail-
ure: role of parathyroid hormone and pulmonary
artery calcifications. Chest 2003;124:2093–7.
6 Bloembergen WE, Port FK, Mauger EA, Wolfe RA.
Comparison of mortality between patients treated
with hemodialysis and peritoneal dialysis. J Am Soc
Nephrol 1995;6:177–83.
7 Tarrass F, Benjelloun M, Medkouri G, Hachim K,
Benghanem MG, Ramdani B. Doppler
echocardiograph evaluation of pulmonary hyperten-
sion in patients undergoing hemodialysis. Hemodial
Int 2006;10:356–9.
8 Goel N, Mittman N, Bhasin A, et al. Pulmonary hy-
pertension in dialysis patients (abstract). J Am Soc
Nephrol 2004;15:323a.
9 Llach F, Massry SG. On the mechanism of secondary
hyperparathyroidism in moderate renal insufficiency.
J Clin Endocrinol Metab 1985;61:601–6.
10 Rahman MH, Hossain MM, Sultana S, Jamal CY,
Karim MA. Correlation of serum parathormone level
with biochemical parameters in chronic renal failure.
Indian Pediatr 2005;42:250–4.
11 Akmal M, Barndt RR, Ansari AN, Mohler JG, Massry
SG. Excess PTH in CRF induces pulmonary calcifi-
cation, pulmonary hypertension and right ventricular
hypertrophy. Kidney Int 1995;47:158–63.
12 Kumbar LM, Mittman N, Sharath C, et al. Pulmonary
hypertension (PHT) is prevalent in chronic kidney
disease prior to initiation of renal replacement
therapy (abstract). J Am Soc Nephrol 2006;17:371a.
13 McQuillan BM, Picard MH, Leavitt M, Weyman AE.
Clinical correlates and reference intervals for pulmo-
nary artery systolic pressure among
echocardiographically normal subjects. Circulation
2001;104:2797–802.
14 Yigla M, Dabbah S, Azzam ZS, Rubin AH, Reisner
SA. Background diseases in 671 patients with moder-
ate to severe pulmonary hypertension. IMAJ
2000;2:684–9.
15 Kato GJ, McGowan V, Machado RF, et al. Lactate
dehydrogenase as a biomarker of hemolysis-associ-
ated nitric oxide resistance, priapism, leg ulcer-
ation, pulmonary hypertension, and death in
patients with sickle cell disease. Blood
2006;107:2279–85.
Corresponding author:
Paul A. Fein, MD, Avram Division of Nephrology, The
Long Island College Hospital, 339 Hicks Street,
Brooklyn, New York 11201 U.S.A.
E-mail:
pafmd@juno.com