Download Relaxin Is an Independent Risk Factor Predicting Death in Male

Relaxin Is an Independent Risk Factor Predicting Death in
Male Patients With End-Stage Kidney Disease
Berthold Hocher, MD; Reinhard Ziebig, PhD; Rolfdieter Krause, MD; Gernot Asmus, MD;
Hans-H. Neumayer, MD; Lutz Liefeldt, MD; Johannes-Peter Stasch, PhD
Background—Patients with end-stage kidney disease (ESKD) have a reduced life expectancy mainly as the result of
cardiovascular diseases. Relaxin has been implicated in the pathogenesis of cardiovascular diseases. We analyzed the
impact of relaxin on death in patients with ESKD.
Methods and Results—Patients (n ⫽245; 122 women, 123 men) on long-term hemodialysis were followed for 1140 days
for death. Blood samples for analysis of relaxin, C-reactive protein, Troponin T, cholesterol, HDL, brain natriuretic
peptide, and albumin were taken at study entry. Survival was compared by the Kaplan-Meier method and Cox regression
analysis. One hundred seven patients died during the observation period; 66 died of cardiovascular diseases and 28 died
of infectious diseases. Elevated serum relaxin concentrations (greater than median) predicted death in male but not in
female patients with ESKD: All-cause death (men: relative risk, 2.63; 95% CI, 1.34 to 5.12; P⫽0.005; women: relative
risk, 0.671; 95% CI, 0.33 to 1.35; P⫽0.262) and cardiovascular death (men: relative risk, 2.95; 95% CI, 1.20 to 7.21;
P⫽0.018; women: relative risk, 0.639; 95% CI, 0.26 to 1.56; P⫽0.324).
Conclusions—Relaxin is an independent risk factor predicting death in male patients with ESKD on chronic hemodialysis.
(Circulation. 2004;109:2266-2268.)
Key Words: cardiovascular diseases 䡲 kidney 䡲 risk factors 䡲 mortality
I
njections of serum from pregnant rabbits induced relaxation of the interpubic ligament of guinea pigs.1 The
hormone responsible for this action was named relaxin.
Corpus luteum, decidua, and placenta secrete this hormone
during pregnancy. Relaxin has many important roles in
pregnancy, including softening effects on connective tissue,
reducing uterine contractility, and control of mammary gland
growth and differentiation.1
Recently, however, it was recognized that relaxin also
plays a role in the cardiovascular system. Patients with
chronic heart failure have increased myocardial relaxin gene
expression and elevated plasma relaxin concentrations.2 Relaxin stimulates cardiac ANP secretion3 and increases coronary blood flow through a nitric oxide–mediated mechanism.3
It was furthermore shown that relaxin is a vasodilator of small
systemic resistance arteries.4 Relaxin is also involved in the
regulation of cardiac5 and renal6 collagen synthesis.
We analyzed whether relaxin is a mortality risk factor in
patients with end-stage kidney disease (ESKD). Patients with
ESKD have a substantially reduced life expectancy. The
mortality risk is about 50% higher in elderly patients with
ESKD and even much more in younger patients with ESDK
as compared with age-matched persons without renal
disease.7,8
Methods
The 245 patients (122 women, 123 men) were recruited from two
dialysis centers (KfH Dialysezentrum-Neukölln, Berlin, and KfH
Dialysezentrum-Moabit, Berlin). The patients are dialyzed at least 3
times weekly in 4- to 6-hour sessions, sufficient to provide an urea
kinetic on dialysis (KT/V) of at least 1.3. The included patients
represent all patients being on stable hemodialysis in these centers
without actual health problems. All patients were included in the
study in March 2000 and followed for 1140 days. The following
patients characteristics were documented: age, gender, body mass
index, cause of ESKD, time on dialysis, diabetes, hypertension, and
coronary heart disease (patients with a history of myocardial infarction, coronary heart disease confirmed by heart catheterization, or
typical stable angina).
Time points and cause of death were documented. Blood samples
were taken before start of hemodialysis at study entry. The study was
approved by the local ethics committee.
Albumin, cholesterol, HDL, C-reactive protein (CrP), and Troponin T (TnT) were measured by standardized methods.9 Relaxin and
brain naturetic peptide (BNP) concentrations were analyzed with the
use of a commercial ELISA (Immundiagostik GmbH, Bensheim,
Germany), according to the manufacturer’s instructions.
Considering the gender-dependent synthesis of relaxin,1 KaplanMeier analysis and Cox regression analysis were done for women
and men separately. We included relaxin a priori and factors known
to have an influence on the end point of death in patients with ESKD
(age, time on dialysis, diabetes, preexisting coronary heart disease,
TnT, CrP, albumin, total cholesterol, HDL cholesterol, and BNP).
Received March 4, 2004; de novo received February 11, 2004; revision received March 25, 2004; accepted March 31, 2004
From the Center for Cardiovascular Research, Departments of Nephrology (B.H., H.-H.N., L.L.) and Laboratory Medicine (R.Z.), Charité, Berlin,
Germany; KfH Dialysezentrum, Moabit, Berlin, Germany (R.K.); KfH Dialysezentrum Sonnenallee, Berlin, Germany (G.A.); and Institute of
Cardiovascular Research (J.-P.S.), Bayer AG, Wuppertal, Germany.
Correspondence to Priv Doz Dr Berthold Hocher, Humboldt University of Berlin, University Hospital Charité, Center for Cardiovascular Research,
Hessischestr 3-4, D-10115 Berlin, Germany. E-mail [email protected]
© 2004 American Heart Association, Inc.
Circulation is available at http://www.circulationaha.org
DOI: 10.1161/01.CIR.0000128598.72920.B5
2266
Hocher et al
TABLE 1.
Clinical Data and Serum Parameters at Study Entry
N
245
Age
63.5⫾5.8
Gender, male/female, n
123/122
Diabetes mellitus
84
Coronary heart disease
157
Hypertension
220
Smoker
101
Time on dialysis, y
5.0⫾4.8
Body mass index, kg/m2
25.2⫾4.6
Serum albumin, g/dL
3.74⫾0.34
CrP, mg/dL
1.26⫾1.30
Relaxin Predicts Death in Men
2267
ESKD. Calculating the risk increase per additional 5 pg/mL
relaxin, using the same cofactors as in the initial Cox regressions
(age, time on dialysis, diabetes, preexisting coronary heart
disease, TnT, CrP, albumin, total cholesterol, HDL cholesterol
and BNP), revealed again that relaxin is an all-cause mortality
risk factor for men (all-cause mortality risk per 5 pg/mL relaxin
for men: 1.077; 95% CI, 1.015 to 1.1143; P⫽0.014; all-cause
mortality risk per 5 pg/mL relaxin for women: 1.032; 95% CI,
0.962 to 1.103; P⫽0.350). For cardiovascular death of men, there
was a trend for increased mortality risk per 5 pg/mL relaxin steps
(P⫽0.08). In female patients with ESKD, relaxin had no significant
impact (P⫽0.66) on cardiovascular death. This latter analysis
suggests that there is a linear relation between mortality risk and
increasing relaxin values for men at least for all-cause death.
TnT, ␮g/L
0.125⫾0.449
Total cholesterol, mg/dL
203.2⫾57.27
Discussion
HDL cholesterol, mg/dL
42.5⫾12.3
BNP, pg/mL
14.4⫾35.4
Relaxin, pg/mL
33.8⫾39.3
This study demonstrates that elevated plasma relaxin concentrations are independently associated with all-cause and
cardiovascular death in male but not female patients with
ESKD.
In an earlier study,9 we demonstrated that risk factors
predicting death in our cohort such as diabetes, elevated TnT,
and low albumin are similar to those detected in other
studies,7,8,10 indicating that our present findings are of general
impact in patients with ESKD.
Our finding of a gender-dependent association of elevated
plasma relaxin concentrations with death is based on several
independent statistical approaches: (1) Kaplan-Meier survival
curves showed only an impact of relaxin on death in male
patients with ESKD. (2) Two independent sets of Cox
regressions revealed that relaxin has a gender-dependent
impact on death in patients with ESKD. We first performed
Cox regression analysis by using the median of all measured
relaxin values as cutoff. The principally similar finding was
seen in Cox regressions using the mortality risk increase per
additional 5 pg/mL relaxin.
It is also noteworthy that only male relaxin knockout mice
have development of cardiac alterations (increased left ventricular collagen synthesis and impaired left ventricular function).5 The molecular pathways explaining the gender dependency remains to be clarified. Relaxin effects are—at least
partially—mediated by an increased NO synthesis.3 Considering that NO synthesis is higher in female subjects,11,12
potential cardiovascular relaxin effects might be masked in
female patients with ESKD. We furthermore suggest that
elevated plasma relaxin concentrations in patients with
ESKD, both male and female, could be a compensatory
phenomenon to counteract developing cardiac dysfunction
and coronary artery disease. Female patients probably are
more sensitive to relaxin and could actually benefit from
relaxin’s cardiovascular actions, whereas male patients could
be less sensitive than female patients to relaxin and have little
or no cardiac protection from relaxin. In support of that
notion (compensatory increased synthesis of relaxin in the
course of chronic heart damage in uremic patients with
ESKD) is the observation that plasma relaxin and left
ventricular relaxin mRNA are elevated in patient with chronic
heart failure.2 Adding BNP, a parameter that describes heart
failure in patients on hemodialysis,13 to the Cox regression
All continuous parameters are given as mean⫾SD.
Cutoff values for known risk factors were determined as recently
described.9 The relaxin cutoff value was calculated by using the
median of all relaxin values. We additionally used relaxin as a
continuous variable in a separate series of Cox regressions. We
calculated in this series the mortality risk increase per additional 5
pg/mL relaxin. All data were analyzed with the use of SPSS for
Windows, Version 11.5.
Results
The underlying renal diseases are as follows: Sixty-five of the
245 patients had diabetic nephropathy, 38 patients had
hypertensive nephrosclerosis, 30 patients had chronic glomerulonephritis, 28 patients had autosomal dominant polycystic
kidney disease, and 20 patients had analgesic nephropathy.
The remainder had various rare kidney diseases. In 40
patients, the underlying renal disease was unknown. The
patient characteristics are given in Table 1. No patient was
lost to follow-up. One hundred seven patients died; in 70
cases, patients died of cardiovascular disease, including 12
with acute myocardial infarction, 18 of sudden cardiac death,
and 29 of chronic heart failure. The highest relaxin concentrations were seen in patients who died later of myocardial
infarction. However, the differences in this subgroup as
compared with survivors were not significant, most probably
because of the low number of cases. (data not shown).
Infectious diseases were the second most common cause of
death (27 cases). Five patients died of cancer. The 5 remaining death cases were related to accidents, hyperkalemia, and
in 1 case the cause of death was unknown.
The median plasma relaxin concentration was 28.8 pg/mL
(range, 0 to 240 pg/mL) in the 245 patients with ESKD.
Kaplan-Meier survival curves indicate that relaxin predicts
all-cause and cardiovascular death in male but not female
patients with ESKD (data not shown).
Cox regressions performed with the median of all measured relaxin values in the entire study population as a cutoff
for relaxin (Tables 2) revealed that relaxin is a predictor of
all-cause as well as cardiovascular death in male patients with
2268
Circulation
May 18, 2004
TABLE 2. Cox Proportional Hazards Analysis of Factors Predicting All-Cause and Cardiovascular Death in
Male and Female Patients With ESKD
Men
Women
Relative Risk
95% CI
P
Relative Risk
95% CI
P
Coronary heart disease, yes
1.79
0.80–4.02
0.159
2.87
1.25–6.64
0.014
Diabetes, yes
1.53
0.80–2.92
0.200
2.12
0.97–4.67
0.062
Age (⬎64.5 y)
1.73
0.91–3.30
0.094
1.46
0.65–3.30
0.364
Time on dialysis (⬎5.5 y)
1.10
0.55–2.18
0.795
1.97
0.96–4.03
0.064
Troponin T (⬎0.039 ␮g/L)
9.06
2.62–31.35
0.001
3.22
1.60–6.49
0.001
CrP (⬎0.555 mg/dL)
1.46
0.70–3.04
0.308
1.24
0.59–2.62
0.566
Albumin (⬍3.71 g/dL)
1.48
0.78–2.82
0.233
1.79
0.81–3.95
0.154
Cholesterol (⬍189.5 mg/dL)
1.10
0.59–2.07
0.779
1.30
0.65–2.62
0.459
HDL cholesterol (⬎39.1 mg/dL)
0.85
0.45–1.57
0.595
0.76
0.35–1.63
0.482
BNP (⬎2.5 pg/mL)
2.74
1.43–5.22
0.002
0.910
0,462–1.793
0.786
Relaxin (⬎28.8 pg/mL)
2.63
1.34–5.12
0.005
0.671
0.33–1.35
0.262
Coronary heart disease, yes
3.40
0.93–12.42
0.064
4.08
1.21–13.77
0.023
Diabetes, yes
2.57
1.04–6.33
0.040
1.83
0.67–5.03
0.237
Age (⬎64.5 y)
2.09
0.88–4.97
0.093
0.92
0.34–2.51
0.871
Time on dialysis (⬎5.5 y)
1.31
0.51–3.34
0.572
0.91
⫺0.33–2.51
0.849
Troponin T (⬎0.039 ␮g/L)
17.17
All-cause mortality
Cardiovascular mortality
2.13–138.29
0.008
4.20
1.60–11.07
CrP (⬎0.555 mg/dL)
1.60
0.57–4.49
0.374
1.31
0.51–3.38
0.576
Albumin (⬍3.71 g/dL)
1.43
0.581–3.51
0.438
1.56
0.57–4.26
0.385
Cholesterol (⬍189.5 mg/dL)
1.16
0.51–2.62
0.723
1.41
0.56–3.51
0.464
HDL cholesterol (⬎39.1 mg/dL)
0.82
0,35–1.91
0.651
0.57
0.22–1.48
0.247
BNP (⬎2.5 pg/mL)
4.44
1.73–11.34
0.002
0.75
0.31–1.82
0.526
Relaxin (⬎28.8 pg/mL)
2.95
1.20–7.21
0.018
0.639
0.26–1.56
0.324
models has no major impact on relaxin mortality risk factors
or probability values. This suggests that relaxin is not mainly
reflecting heart failure in male patients with ESKD. This
hypothesis is supported by the finding that there was a trend
toward highest relaxin concentrations in patients who died
later of myocardial infarction but not of heart failure.
The present study also indicated that BNP is a predictor of
death only in male patients with ESKD. The underlying molecular pathways are again unknown. However, it is of note that in
the general population, elevated plasma BNP was associated
with a greater likelihood of future blood pressure increase only
in men.14 In addition, Luchner et al15 showed that plasma BNP
concentrations are much better correlated to left ventricular mass
and function in men as compared with women.
Acknowledgments
This study was partially supported by grant DFG Ho1665/5-2.
References
1. Bani D. Relaxin: a pleiotropic hormone. Gen Pharmacol. 1997;28:13–22.
2. Dschietzig T, Richter C, Bartsch C, et al. The pregnancy hormone relaxin
is a player in human heart failure. FASEB J. 2001;15:2187–2195.
3. Samuel CS, Parry LJ, Summers RJ. Physiological or pathological: a role for
relaxin in the cardiovascular system? Curr Opin Pharmacol. 2003;3:152–158.
4. Fisher C, MacLean M, Morecroft I, et al. Is the pregnancy hormone
relaxin also a vasodilator peptide secreted by the heart? Circulation.
2002;106:292–295.
0.004
5. Du XJ, Samuel CS, Gao XM, et al. Increased myocardial collagen and
ventricular diastolic dysfunction in relaxin deficient mice: a genderspecific phenotype. Cardiovasc Res. 2003;57:395– 404.
6. Garber SL, Mirochnik Y, Brecklin CS, et al. Relaxin decreases renal
interstitial fibrosis and slows progression of renal disease. Kidney Int.
2001;59:876 – 882.
7. Port FK. Morbidity and mortality in dialysis patients. Kidney Int. 1994;
46:1728 –1737.
8. Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis. 1998;32(5 suppl
3):S112–S119.
9. Hocher B, Ziebig R, Altermannb C, et al. Different impact of biomarkers
as mortality predictors in diabetic and non-diabetic patients on hemodialysis. J Am Soc Nephrol. 2003;14:2329 –2337.
10. Dierkes J, Domrose U, Westphal S, et al. Cardiac troponin T predicts
mortality in patients with end-stage renal disease. Circulation. 2000;102:
1964 –1969.
11. Kauser K, Rubanyi GM. Gender difference in bioassayable endotheliumderived nitric oxide from isolated rat aortae. Am J Physiol. 1994;267:
H2311–H2317.
12. Knot HJ, Lounsbury KM, Brayden JE, Nelson MT. Gender differences in
coronary artery diameter reflect changes in both endothelial Ca2⫹ and
eNOS activity. Am J Physiol. 1999;276:H961–H969.
13. Zoccali C, Mallamaci F, Benedetto FA, et al. Cardiac natriuretic peptides
are related to left ventricular mass and function and predict mortality in
dialysis patients. J Am Soc Nephrol. 2001;12:1508 –1515.
14. Freitag MH, Larson MG, Levy D, et al. Plasma brain natriuretic peptide
levels and blood pressure tracking in the Framingham Heart Study.
Hypertension. 2003;41:978 –983.
15. Luchner A, Brockel U, Muscholl M, et al. Gender-specific differences of
cardiac remodeling in subjects with left ventricular dysfunction: a
population-based study. Cardiovasc Res. 2002;53:720 –727.