Differential association of proinflammatory cytokines with left ventricular diastolic dysfunction in subjects with and without continuous ambulatory peritoneal dialysis

Nutrition, Metabolism & Cardiovascular Diseases (2012) 22, 974e980 available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/nmcd

Differential association of proinflammatory cytokines with left ventricular diastolic dysfunction in subjects with and without continuous ambulatory peritoneal dialysis J.-K. Lee b,c,1, H.-H. Lin a,1, C.-T. Tsai a, J.-J. Chen a,d, C.-C. Kuo d,e, Y.-C. Lien e, J.-W. Lin a,d,*, J.-W. Huang e, S.-W. Hwang a, J.-J. Hwang a, C.-D. Tseng a, F.-T. Chiang a,b, J.-J. Chen a, C.-K. Wu a,d,f,** a

Division of Cardiology, Department of Internal Medicine, National Taiwan University College of Medicine and Hospital, Taipei, Taiwan b Department of Laboratory Medicine, National Taiwan University Hospital, Taipei, Taiwan c Graduate Institute of Biomedical Electronics and Bioinformatics, College of Electrical Engineering & Computer Science, National Taiwan University, Taipei, Taiwan d Department of Internal Medicine, National Taiwan University College of Medicine and Hospital Yun-Lin Branch, Yun-Lin, Taiwan e Division of Nephrology, Department of Internal Medicine, National Taiwan University College of Medicine and Hospital, Taipei, Taiwan f Graduate Institute of Clinical Medicine, College of Medicine, National Taiwan University, Taipei, Taiwan Received 28 July 2010; received in revised form 29 November 2010; accepted 7 January 2011 Available online 17 May 2011

KEYWORDS Continuous ambulatory peritoneal dialysis; Left ventricular diastolic dysfunction; Inflammation

Abstract Background and aims: The association between inflammation and left ventricular (LV) diastolic dysfunction in continuous ambulatory peritoneal dialysis (CAPD) and non-CAPD patients is not established. The objective of this study was to test the above association and whether inflammation interacts with CAPD to increase LV diastolic dysfunction risks. Methods and results: 120 subjects with normal creatinine levels and 101 CAPD patients were recruited. Echocardiographic parameters were assessed in all patients. The participants were classified as having LV diastolic dysfunction by echocardiographic findings including mitral inflow E/A ratio < 1, deceleration time > 220 cm/s, or decreased peak annular early diastolic

* Corresponding author. Department of Internal Medicine, National Taiwan University College of Medicine and Hospital Yun-Lin Branch, Cardiovascular Center, 579 Yun-Lin Road Section 2, Dou-Liou City, Yun-Lin 640, Taiwan. Tel.: þ886 922861953; fax: þ886 55373257. ** Corresponding author. Department of Internal Medicine, National Taiwan University College of Medicine and Hospital Yun-Lin Branch, Cardiovascular Center, 579 Yun-Lin Road Section 2, Dou-Liou City, Yun-Lin 640, Taiwan. Tel.: þ886 972655370; fax: þ886 72221195. E-mail addresses: [email protected] (J.-W. Lin), [email protected] (C.-K. Wu). 1 The first two authors contributed equally to this work. 0939-4753/$ - see front matter ª 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.numecd.2011.01.001

Differential association of proinflammatory cytokines

975

velocity in tissue Doppler imaging. Blood was sampled at the baseline for measurement of inflammation markers, including tissue necrosis factor-alpha (TNF-a) and interleukin-6 (IL-6). Subjects with LV diastolic dysfunction had higher proinflammation cytokines levels in both groups. Inflamed markers correlated significantly with echocardiography parameters for LV diastolic dysfunction in patients receiving CAPD. In a multivariate regression analysis adjusting for all the factors associated with LV diastolic dysfunction, inflammation is still significantly associated with left ventricular diastolic dysfunction (TNF-alpha, OR: 2.6, 95% CI: 2.0e3.35, p < 0.001; IL-6, OR: 1.26, 95% CI: 1.25e1.26, p Z 0.01). In addition, the interaction of CAPD and inflammation significantly contributed to the development of LV diastolic dysfunction (CAPD* TNF-a: OR: 1.45, 95% CI: 1.13e1.79, P Z 0.004). Conclusion: We found inflammation plays a vital role for LV diastolic dysfunction especially in CAPD patients. A synergistic effect between CAPD and inflammation, especially TNF-a, would further aggravate LV diastolic dysfunction. ª 2011 Elsevier B.V. All rights reserved.

Introduction According to cross-sectional and population-based studies, approximately 50% of patients with heart failure (HF) have a normal or near normal ejection fraction and, therefore, are referred to as having diastolic heart failure or heart failure with normal ejection fraction (HFNEF) [1e3]. Also, other observational studies have shown that the prognosis of patients with HFNEF appears to be similar to that of patients with HF and reduced LV ejection fraction [4,5]. Until now there has been much controversy about the pathophysiology of HFNEF. There are several mechanisms associated with LV diastolic dysfunction proposed, including impaired LV relaxation, increased LV passive stiffness, pericardial and endocardial disease, impaired neurohormonal regulation and even genetic factors [6e8]. However, so far, the mechanism of cardiac diastolic dysfunction is not fully understood. Immunoinflammatory activation has been demonstrated to play a pivotal role in the development and in the progression of heart failure by some studies [9e11]. Through influencing cardiac contractility, inducing hypertrophy, and promoting myocyte apoptosis and fibrosis, it has been suggested that inflamed cytokines contribute to left ventricular remodeling and to the progression of heart failure [10,12e14,39]. Besides, circulating levels of tumor necrosis factor-a (TNF-a) and interleukin-6 (IL-6) have been found to be elevated in direct relation to deteriorating functional class of heart failure, predicting a worse outcome. Such inflamed cytokines could promote and maintain inflammation locally and stimulate fibroblasts to produce collagen and subsequently lead to increased myocardial stiffness and diastolic dysfunction [10]. The association between plasma levels of TNF-a and IL-6 and diastolic heart failure in some specific groups, such as patients with coronary artery disease or patients with newly diagnosed systolic heart failure has been suggested recently [15,16]. In patients with renal failure, it is now widely accepted that there is a high prevalence of inflammation and oxidative stress, both of which are associated with the high rate of cardiovascular events [17e19]. Similar findings are reported by several groups in both hemodialysis (HD) and continuous ambulatory peritoneal dialysis (CAPD) patients

alike [20]. In addition to inflammation, patients with renal failure suffered more for fluid overload, a high prevalence of hypertension and LV hypertrophy (which is a physiological response to pressure and volume overload) which all, it is suggested, contribute to the higher prevalence of LV diastolic dysfunction [21,22]. In CAPD patients, an association between inflammation and fluid overload has been suggested [23,24]. However, there is no published study on the relationship between inflammation and LV diastolic dysfunction in CAPD patients and whether anti-inflammation medication (e.g. statin) has a specific role in this group of patients remains to be determined. Tissue Doppler imaging (TDI) has been a simple, reproducible and widely available noninvasive tool for the assessment left ventricular diastolic function. In 2007, the European Working Group on HFNEF proposed a new diagnostic algorithm which implemented TDI techniques as the basic tool to help diagnose LV diastolic dysfunction and HFNEF [25]. Therefore, this study is designed to explore the possible association between LV diastolic dysfunction and inflammation reflected by serum IL-6 and tissue TNF-a levels in CAPD patients and subjects with normal creatinine. We would also investigate whether there is a synergistic effect for inflammation and CAPD for LV diastolic dysfunction.

Methods Study subjects We had two study groups. The first group (group 1) consisted 120 subjects with normal renal function (serum creatinine <1.5 mmol) recruited sequentially from the outpatient clinic or the cardiovascular ward of National Taiwan University Hospital and its affiliates from July 2007 to March 2009. The second group (group 2) enrolled 101 patients consecutively that had received CAPD for more than six months at the same hospital. These participants received conventional glucose-based lactate-buffered CAPD solutions (UltraBag; Baxter Healthcare SA, Singapore). Patients who had hepatic disease, history of myocardial infarction, coronary intervention, cardiac myopathy, pericardial disease, significant valvular heart disease (moderate), chronic obstructive pulmonary disease, chronic atrial fibrillation,

976 clinical signs of acute infection or other chronic inflammatory conditions were excluded. Patients who received statins, lipid-lowering agents and other medication that may influence plasma inflammatory cytokine levels were excluded also. All the subjects underwent echocardiographic examinations as well as blood sampling for the estimation of plasma TNF-a and IL-6. Written informed consent was obtained from each individual and the study was reviewed and approved by the Institutional Review Board.

Demographic and clinical data Demographic data were collected from medical records. Hypertension was defined as a systolic blood pressure of 140 mm Hg, diastolic blood pressure of 90 mm Hg, or the use of at least one class of antihypertensive agent. Type 2 diabetes mellitus was defined as a fasting blood glucose concentration >126 mg/dL and/or the use of at least one oral hypoglycemic agent. Medication information, such as the use of angiotensin-converting enzyme (ACE) inhibitors and/or ARBs, calcium channel blockers, diuretics, nitrates, or b-blockers for at least 6 months before the examination of echocardiography was recorded. Serum cholesterol, triglycerides, low density lipids, high density lipids and white blood cells were measured within one week of echocardiographic examination.

Measurements of plasma TNF-a and IL-6 In all the patients, blood samples were collected from the antecubital vein between 08:00 and 10:00 h, in supine position after 12 h of fasting. The blood was spun at 2000g for 15 min; then the plasma was separated and stored at 80  C until used. Interleukin-6 was measured with high-sensitivity enzyme linked immunoassay (ELISA) (catalog no. HST600B, R & D Systems, Minneapolis, MN, USA) with the minimal detectable dose of a range from 0.016 to 0.110 pg/ml. We used the ELISA method for the quantitative determination of human TNF-a in duplicate in serum samples of the participants by means of the relevant kit Quantikine HS/human TNF-a immunoassay (catalog no.HSTA00D, R & D Systems, Minneapolis, MN, USA). The minimal detectable dose ranged from 0.038 to 0.191 pg/ml. The calculated overall interassay and intra-assay coefficients of variation were: 6.5e9.6% and 6.9e7.8% for IL-6; 7.4e10.6% and 3.1e8.5% for TNF-a respectively.

Echocardiographic evaluation LV diastolic dysfunction was defined according to our previous work and the consensus from the European Society of Cardiology. Briefly as 1) normal systolic function (ejection fraction  50%), and 2) echocardiographic evidence of LV diastolic dysfunction: a mitral inflow E/A ratio < 1, deceleration time > 220 cm/s, or decreased peak annular early diastolic velocity of the lateral mitral annulus in tissue Doppler imaging (e’) [E/e’  15] [26,27]. Eighty participants in group 1 were classified as having LV diastolic dysfunction and 40 subjects served as controls and had no symptoms of heart failure and no objective evidence of diastolic dysfunction under echocardiography. In group 2, 35 patients

J.-K. Lee et al. were classified as having LV diastolic dysfunction; the other 60 patients were defined as non-LV diastolic dysfunction.

Statistical analysis Statistical analysis was performed using standard statistical software, SPSS 15.0 software (SPSS, Chicago, IL, USA). Continuous variables are presented as mean values  standard deviation, while categorical variables are presented as frequencies. Associations between categorical variables were tested by the use of Pearson’s chi-squared test. Comparisons between groups of patients on normally distributed continuous variables were performed using the Student’s t-test. Associations between cytokines and Doppler parameters correlated using Pearson’s correlation coefficient. We calculated the significance of the difference between the correlation coefficient of normal renal function and CAPD groups by using the Fisher r-to-z transformation (http://faculty.vassar.edu/lowry/rdiff.html). To determine the adjusted association of diastolic dysfunction and possible risk factors, we used four logistic regression models. Confounders were selected with either clinical apparent factors for LV diastolic dysfunction (such as age, gender, BMI, DM, hypertension) or the exposures with significant association with LV diastolic dysfunction (P < 0.10) after correlation tests (such as TNF-a/IL-6 and interaction to PD in each groups). In model 1, we evaluated the effects of TNF-a and CAPD for LV diastolic dysfunction. In model 2, we evaluated the association of IL-6 and CAPD for LV diastolic dysfunction. To examine the synergistic effects between inflammation and CAPD, we developed interaction terms of CAPD) TNFaand CAPD) IL-6. For patients who received CAPD, the value of CAPD) TNF-a and CAPD) IL-6 would be 1)TNF-a and 1) IL6 whereas in group 1, the value would be zero. Therefore, in model 3, we incorporated CAPD) TNF-a, TNF-a, CAPD as independent variables and in model 4 we incorporated CAPD) IL-6, IL-6, CAPD. All models were adjusted by age, gender, BMI, hypertension and DM. We also incorporated four multiple linear regression models to determine the adjusted association of mitral inflow E/A ratio (as continuous variable) and possible risk factors. A p value less than 0.05 was considered statistically significant.

Results Baseline characteristics The baseline characteristics of two groups of study subjects are shown in Table 1. The CAPD subjects tend to be younger, with higher percentages of hypertension, and with higher inflammation cytokines levels than the normal creatinine function group as found in previous reports.

Plasma levels of TNF-a and IL-6 Bivariate Pearson correlations of LV diastolic function parameters and plasma levels of cytokines are shown in Table 2a and Table 2b. In subjects with normal creatinine levels, TNF-a had only a mildly significant association with mitral valve inflow deceleration time (DT) but had no correlation to other LV diastolic function parameters (Table 2a). In the

Differential association of proinflammatory cytokines

977

Table 1 Baseline characteristics, echocardiographic parameters and cytokines levels in group 1 (normal creatinine levels) and group 2 (CAPD) subjects.

Age (years) Sex (M/F) BMI (kg/m2) Diabetes mellitus (%) Hypertension (%) Smoking (%) WBC Cholesterol Triglyceride Low density cholesterol Anti-hypertension therapy ACEI þ ARB (%) b-blocker (%) Calcium channel blocker (%) Echocardiographic Data LA diameter (mm) LVEF (%) LVEDD (mm) LVESD (mm) Cytokine levels Interleukin-6 (pg/ml) Tissue necrosis factor e alpha (pg/ml)

Group1 (n Z 120)

Group 2 (n Z 101)

72.3  9.6 46/75 24.8  3.8 24.8

68.7  8.9 42/59 23.5  3.1 19.8

71.1 32.2 7179 197.9 138.0 126.7

79.2 24.8 7406 200.6 153.6 115.3

   

2030 37.4 94.8 37.3

28.1 34.7 38.0

35.4 66.3 46.1 28.7

   

2039 51.8 109.3 33.6

34.7 41.6 47.5

   

6.1 8.4 6.1 5.8

36.0 60.7 49.8 31.1

significantly associated with all LV diastolic dysfunction parameters (E/A, e’, E/e’, DT) (Table 2b). Figure 1A and B displayed bivariate correlations of E/A and plasma levels of cytokines in both groups. Using the Fisher r-to-z transformation, we calculated a z value to assess the significance of the difference between two correlation coefficients. We found the correlation between LV diastolic function parameters and inflammatory markers had significant difference between group 1 and group 2. Patients receiving CAPD had significantly higher correlation than the normal creatinine group for TNF-a VS E/A (z Z 2.24, p Z 0.01), TNF-a VS E/e’ (z Z 1.74, p Z 0.04) and TNF-a VS DT (z Z 5, p < 0.001) (Table 2a). Similarly, the CAPD group had a higher correlation in IL-6 VS E/A (z Z 1.85, p Z 0.03), IL-6 VS e’ (z Z 2.63, p Z 0.004), IL-6 VS E/e’ (z Z 4.05, p < 0.001) and IL-6 VS DT (z Z 2.15, p Z 0.02) (Table 2b). By considering the E/A ratio as continuous variable, the multiple linear regression model adjusting for potential confounders still revealed that the interaction between PD and TNF-a are associated with changes of E/A ratio (B Z 0.023, p Z 0.019) (Fig. 1A and Supplementary Table 1).

Inflammation for LV diastolic dysfunction

   

5.9 10.7 7.2 7.3

3.40  2.88

37.58  7.85

4.78  2.24

13.5  3.5

LV, left ventricle; BMI, body mass index; WBC, white blood count; ACEI, angiotensin-converting enzyme blocker; ARB, angiotensin II type I receptor blocker; LA, left atrium; LVEF, left ventricular ejection fraction; LVEDD, left ventricular end diastolic dimension; LVESD, left ventricular end systolic dimension.

same way, IL-6 was only weakly associated with e’ but not with any other diastolic function parameters (Table 2b). On the other hand, in group 2 (patients receiving CAPD), TNFa had significant weak to moderate association with mitral inflow E/A, E/e’ and DT (Table 2a) whereas IL-6 are

The proinflammatory cytokines level was higher in individuals with LV diastolic dysfunction in both groups. To adjust for the confounding effect of other clinical factors, we performed logistic multivariate analysis on all participants (n Z 210) and found that proinflammatory cytokines were still associated with LV diastolic dysfunction after adjusting for age, gender, BMI, DM, HTN, BMI, CAPD, IL-6 or TNFa (TNF-a: OR: 2.6, 95% CI: 2.0e3.35, P < 0.001 and IL-6: OR: 1.26, 95% CI: 1.25e1.26, P Z 0.01, respectively; Table 3, model 1 and model 2).

Interaction of CAPD and inflammation on LV diastolic dysfunction We further evaluated the interaction of CAPD and proinflammatory cytokines and the association with LV diastolic dysfunction by multivariate logistic regression analysis. After adjusting the confounding factors and incorporating the interaction terms of CAPD) IL-6 and CAPD) TNF-a into the analysis, the effect of IL-6 and TNF-awere still significant (TNF-a: OR: 1.62, 95% CI: 1.27e2.06, P < 0.001 and IL-6: OR: 1.30, 95% CI: 1.01e1.54, P Z 0.03, respectively; Table 3, model 3 and model 4). Each unit increase

Table 2a Correlation between echocardiographic parameters of left ventricular diastolic function and plasma concentrations of TNF-a in group 1 (normal creatinine levels) and group 2 (CAPD) subjects.

E/A e’ E/e’ DT

Group 1 (n Z 120)

Group 2(n Z 101)

Difference between two correlation coefficients

r

P

r

p

z

p

0.18 0.009 0.05 0.26

0.053 0.9 0.63 0.005

0.453 0.09 0.28 0.73

0.004 0.34 0.02 <0.001

2.24 0.59 1.74 5

0.01 0.28 0.04 <0.001

E, mitral valve ejection flow; A, mitral valve atrium flow; e’, peak mitral annular early diastolic velocity; DT, mitral valve ejection flow deceleration time; r, correlation coefficient; z, value to assess the significance of the difference between two correlation coefficients.

978

J.-K. Lee et al.

Table 2b Correlation between echocardiographic parameters of left ventricular diastolic function and plasma concentrations of IL-6 in group 1 (normal creatinine levels) and group 2 (CAPD) subjects.

E/A e’ E/e’ DT

Group 1 (n Z 120)

Group 2(n Z 101)

Difference between two correlation coefficients

r

p

r

p

z

P

0.1 0.2 0.05 0.117

0.28 0.04 0.6 0.21

0.34 0.51 0.54 0.39

0.001 <0.001 0.02 0.003

1.85 2.63 4.05 2.15

0.03 0.004 <0.001 0.02

E, mitral valve ejection flow; A, mitral valve atrium flow; e’, peak mitral annular early diastolic velocity; DT, mitral valve ejection flow deceleration time; r, correlation coefficient; z, value to assess the significance of the difference between two correlation coefficients.

of TNF-a will increase odds of LV diastolic dysfunction by 1.62 and each unit increase of IL-6 will increase the odds by 1.30. Moreover, we found that the interaction between CAPD and proinflammatory cytokines offered a synergistic effect for the development of LV diastolic dysfunction (CAPD) TNF-a: OR: 1.45, 95% CI: 1.13e1.79, P Z 0.004 and CAPD) IL-6: OR: 1.12, 95% CI: 0.65e1.56, P Z 0.07 respectively; Table 3, model 3 and model 4). Comparing to subjects with normal creatinine levels, each unit elevation of TNF-a will further add a 1.45 chance for the development of LV diastolic dysfunction due to the synergistic effect of TNF-a and CAPD.

Discussion

Figure 1 Correlation between tissue necrosis factor-alpha (TNF-a) and mitral inflow E/A ratio (Fig. 1A; group 1, R Z 0.18, p Z 0.053; group 2, R Z 0.453, p Z 0.004; p for interaction of TNF-a and peritoneal dialysis Z 0.037); interleukin-6 (IL-6) and mitral inflow E/A ratio (Fig. 1B; group 1, R Z 0.1, p Z 0.28; group 2, R Z 0.34, p Z 0.001; p for interaction of IL-6 and peritoneal dialysis Z 0.08).

In the present study, we find there is a significant correlation between inflammatory cytokines and LV diastolic dysfunction in CAPD patients and the correlation was significantly stronger than in subjects with normal creatinine levels. Concurrently, we also notice that there is an interaction between systemic inflammation and CAPD, especially TNF-a and CAPD, which had a synergistic effect for the development of LV diastolic dysfunction in CAPD subjects. To our best knowledge, this is the first study to show that systemic inflammation and the interaction of CAPD and inflammation would deteriorate LV diastolic function in CAPD patients. Cardiovascular disease accounts for premature death in more than 50% of dialysis patients from Europe and North America [28]. Besides, of conventional risk factors (i.e., Framingham risk score), such as hypertension, diabetes mellitus and dyslipidemia, it is now widely acknowledged that inflammation plays a pivotal role in the development of cardiovascular disease in end-stage renal disease (ESRD) patients. Although there is no published report to show definite prevalence rate of diastolic heart failure in CAPD patients, a higher prevalence of LV diastolic heart failure in such subjects than in the general population is expected owing to several possible mechanisms, such as inflammation, fluid overload, hypertension, rennin-angiotensin aldosterone system (RAAS) activation, and left ventricular hypertrophy [6,21,22,29,39]. In our study, we found that systemic inflammation was associated with the development of LV diastolic dysfunction in patients receiving CAPD. For the concomitant higher mortality and morbidity with diastolic heart failure than those without, it is an important factor we should consider prudently before we choose CAPD as a modality of renal replacement therapy.

Differential association of proinflammatory cytokines Table 3 Multivariate logistic regression models for parameters associated with left ventricular diastolic dysfunction. Odds ratio Model 1 TNF-alpha Model 2 IL-6 Model 3 TNF-alpha TNF-alpha  CAPD Model 4 IL-6 IL-6  CAPD

95% Confidence interval

P

2.6

2.0e3.35

<0.001

1.26

1.25e1.26

0.01

1.62 1.45

1.27e2.06 1.13e1.79

<0.001 0.004

1.30 1.12

1.01e1.54 0.65e1.56

0.03 0.07

Abbreviations: IL-6, interleukin-6; TNF-alpha, TNF-alpha. ORs and 95% CIs were estimated by logistic regression. All models were adjusted for age, gender, BMI, DM, hypertension and the main effect of peritoneal dialysis during case selection.

Inflammation has become one of the central issues in the pathogenesis of heart failure over the past decades. So far there have been few published reports that demonstrated the participation of proinflammatory cytokines in the development of LV diastolic dysfunction [30,39]. Recent studies suggest that inflamed cytokines can impact myocardial function via effects on both myocyte contractility and the extracellular matrix [31]. Acting as paracrine, TNF-a and IL-6, which promote and maintain inflammation locally, can stimulate fibroblasts to produce collagen. The latter progressively accumulates, thereby leading to increased myocardial stiffness and diastolic dysfunction [10]. Moreover TNF-a can induce myocardio-depressive effects and cause apoptosis [32]. The development of progressive cardiomyocyte apoptosis plays a critical role on the left ventricular geometry and the adverse cardiac remodeling that occurs in the setting of sustained inflammation. Given all these findings, in ESRD patients receiving CAPD, the prevalence and severity of LV diastolic dysfunction are higher, which is probably secondary to a higher level of serum inflammatory cytokines that means more severe systemic inflammation. Statins (3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors) are a family of lipid-lowering drugs known to exert several effects beyond lipid lowering, including anti-inflammation. As suggested by the JUPITER study, rosuvastatin could lower hs-CRP and reduce cardiovascular events and death. Besides being an effective lipid-lowering agent for CAPD and HD patients with dyslipidemia, several studies have also documented that statin could decrease serum CRP and even IL-6 level in chronic kidney disease (CKD) and dialysis individuals [33e36]. Accordingly, statin is probably a promising therapeutic agent for LV diastolic dysfunction in CAPD patients. But there is still controversy about whether the effects of these could translate into better clinical outcomes. In the German Diabetes and Dialysis (Die Deutsche Diabetes Dialyse Studie or 4D) study [37], death from all causes was similar in the atorvastatin and the placebo groups (RR 0.93, 95% CI 0.79e1.08; P Z 0.33) but atorvastatin therapy resulted in a significant reduction of all cardiac events combined compared with

979 placebo (RR 0.82, 95% CI 0.68e0.99; P Z 0.03). In addition, in the recent AURORA study (A study to evaluate the Use of Rosuvastatin in Subjects on Regular Hemodialysis: An Assessment of Survival and Cardiovascular Events), no significant reduction in cardiovascular mortality and all causes of mortality was shown [38]. Since both of the studies were conducted only in HD patients and since we discovered a specific interaction between CAPD and inflammation for the development of LV diastolic dysfunction, more evidence is needed to define the role of statin therapy in CAPD patients. Our study had some limitations. First, this is an observational study of non-complicated general population and patients under peritoneal dialysis and, as such, other variables that could potentially predict alterations in LV structure and function, such as secondary hypertension, could not be evaluated. Second, both groups of study subjects are still undergoing follow up and it is unclear whether suppression of inflammation could lead to improvement of cardiac diastolic dysfunction. Third, the true association between peritoneal dialysis, inflammation and LV diastolic dysfunction still needs to be verified in animal models or larger prospective randomized control studies.

Conclusion In conclusion, our study shows that there is a significant correlation between LV diastolic dysfunction and serum inflammatory cytokines (TNF-alfa, and IL-6) in patients receiving CAPD. An interaction between CAPD and inflammation, especially TNF-a, was also shown to further aggravate LV diastolic dysfunction. Hence it is reasonable that the association between LV diastolic dysfunction and CAPD was at least in part due to the excess of plasma cytokines, which suggests increase of systemic inflammation.

Conflict of interest There is no conflict of interest.

Acknowledgments This work was in part supported by the IBMS CRC Research Program of Institute of Biomedical Science, Academia Sinica (IBMS-CRC99-P02) and a grant from the National Science Council of R.O.C. (NSC 99-2314-B-002-131-MY3).

Appendix. Supplementary material Supplementary data associated with this article can be found in online version at doi:10.1016/j.numecd.2011. 01.001.

References [1] Hogg K, Swedberg K, McMurray J. Heart failure with preserved left ventricular systolic function: epidemiology, clinical characteristics, and prognosis. JACC 2004;43:317e27. [2] Owan TE, Hodge DO, Herges RM, Jacobsen SJ, Roger VL, Redfield MM. Trends in prevalence and outcome of heart

980

[3] [4]

[5]

[6] [7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17] [18]

[19]

[20]

[21]

J.-K. Lee et al. failure with preserved ejection fraction. N Engl J Med 2006; 355:251e9. Gaasch WH, Zile MR. Left ventricular diastolic dysfunction and diastolic heart failure. Annu Rev Med 2004;55:373e94. Bhatia RS, Tu JV, Lee DS, Austin PC, Fang J, Haouzi A, et al. Outcome of heart failure with preserved ejection fraction in a population-based study. N Engl J Med 2006;355:260e9. Vasan RS, Larson MG, Benjamin EJ, Evans JC, Reiss CK, Levy D. Congestive heart failure in subjects with normal versus reduced left ventricular ejection fraction: prevalence and mortality in a population-based cohort. JACC 1999;33:1948e55. Angeja BG, Grossman W. Evaluation and management of diastolic heart failure. Circulation 2003;107:659e63. Wu CK, Luo JL, Wu XM, Tsai CT, Lin JW, Hwang JJ, et al. A propensity score-based case-control study of renin-angiotensin system gene polymorphisms and diastolic heart failure. Atherosclerosis 2009;205:497e502. Wu CK, Tsai CT, Hwang JJ, Luo JL, Juang JJ-M, Hsu KL, et al. Renin-angiotensin system gene polymorphisms and diastolic heart failure. Eur J Clin Invest 2008;38:789e97. Deswal A, Petersen NJ, Feldman AM, Young JB, White BG, Mann DL. Cytokines and cytokine receptors in advanced heart failure: an analysis of the cytokine database from the Vesnarinone trial (VEST). Circulation 2001;103:2055e9. Sciarretta S, Paneni F, Palano F, Chin D, Tocci G, Rubattu S, et al. Role of the renin-angiotensin-aldosterone system and inflammatory processes in the development and progression of diastolic dysfunction. Clin Sci (Lond) 2009;116:467e77. Ueland T, Aukrust P, Yndestad A, Otterdal K, Froland SS, Dickstein K, et al. Soluble CD40 ligand in acute and chronic heart failure. Eur Heart J 2005;26:1101e7. Finkel MS, Oddis CV, Jacob TD, Watkins SC, Hattler BG, Simmons RL. Negative inotropic effects of cytokines on the heart mediated by nitric oxide. Science 1992;257:387e9. Hirota H, Yoshida K, Kishimoto T, Taga T. Continuous activation of gp130, a signal-transducing receptor component for interleukin 6-related cytokines, causes myocardial hypertrophy in mice. Proc Natl Acad Sci USA 1995;92:4862e6. Yokoyama T, Nakano M, Bednarczyk JL, McIntyre BW, Entman M, Mann DL. Tumor necrosis factor-alpha provokes a hypertrophic growth response in adult cardiac myocytes. Circulation 1997;95:1247e52. Kosmala W, Derzhko R, Przewlocka-Kosmala M, Orda A, Mazurek W. Plasma levels of TNF-alpha, IL-6, and IL-10 and their relationship with left ventricular diastolic function in patients with stable angina pectoris and preserved left ventricular systolic performance. Coron Artery Dis 2008;19:375e82. Chrysohoou C, Pitsavos C, Barbetseas J, Kotroyiannis I, Brili S, Vasiliadou K, et al. Chronic systemic inflammation accompanies impaired ventricular diastolic function, detected by Doppler imaging, in patients with newly diagnosed systolic heart failure (Hellenic Heart Failure Study). Heart Vessel 2009;24:22e6. Stenvinkel P. Inflammation in end-stage renal disease: the hidden enemy. Nephrology 2006;11:36e41 (Review Article). Oberg BP, McMenamin E, Lucas FL, McMonagle E, Morrow J, Ikizler TA, et al. Increased prevalence of oxidant stress and inflammation in patients with moderate to severe chronic kidney disease. Kidney Int 2004;65:1009e16. Arici M, Walls J. End-stage renal disease, atherosclerosis, and cardiovascular mortality: is C-reactive protein the missing link? Kidney Int 2001;59:407e14. Tonbul HZ, Demir M, Altıntepe L, Gu ¨ney I, Yeter E, Tu ¨rk S, et al. Malnutrition-Inflammation-Atherosclerosis (MIA) Syndrome components in hemodialysis and peritoneal dialysis patients. Ren Fail 2006;28:287e94. London GM. Cardiovascular disease in chronic renal failure: pathophysiologic aspects. Semin Dial 2003;16:85e94.

[22] Bajraktari G, Berbatovci-Ukimeraj M, Hajdari A, Ibraimi L, Daullxhiu I, Elezi Y, et al. Predictors of increased left ventricular filling pressure in dialysis patients with preserved left ventricular ejection fraction. Croat Med J 2009;50:543e9. [23] Avila-Diaz M, Ventura MD, Valle D, Vicente-Martinez M, Garcia-Gonzalez Z, Cisneros A, et al. Inflammation and extracellular volume expansion are related to sodium and water removal in patients on peritoneal dialysis. Perit Dial Int 2006; 26:574e80. [24] Gangji AS, Brimble KS, Margetts PJ. Association between markers of inflammation, fibrosis and hypervolemia in peritoneal dialysis patients. Blood Purif 2009;28:354e8. [25] Paulus WJ, Tscho ¨pe C, Sanderson JE, Rusconi C, Flachskampf FA, Rademakers FE, et al. How to diagnose diastolic heart failure: a consensus statement on the diagnosis of heart failure with normal left ventricular ejection fraction by the Heart Failure and Echocardiography Associations of the European Society of Cardiology. Eur Heart J 2007;28:2539e50. [26] Wu CK, Tsai CT, Chang YC, Luo JL, Wang YC, Hwang JJ, et al. Genetic polymorphisms of the angiotensin II type 1 receptor gene and diastolic heart failure. J Hypertens 2009;27:502e7. [27] Wu CK, Luo JL, Tsai CT, Huang YT, Cheng CL, Lee JK, et al. Demonstrating the pharmacogenetic effects of angiotensinconverting enzyme inhibitors on long-term prognosis of diastolic heart failure. Pharmacogenomics J 2010;10:46e53. [28] Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis 1998;32:S112e9. [29] Schiffrin EL, Lipman ML, Mann JFE. Chronic kidney disease: effects on the cardiovascular system. Circulation 2007;116: 85e97. [30] Dinh W, Futh R, Nickl W, Krahn T, Ellinghaus P, Scheffold T, et al. Elevated plasma levels of TNF-alpha and Interleukin-6 in patients with diastolic dysfunction and glucose metabolism disorders. Cardiovasc Diabetol 2009;8:58e64. [31] Prabhu SD. Cytokine-Induced Modulation of cardiac function. Circ Res 2004;95:1140e53. [32] Ing DJ, Zang J, Dzau VJ, Webster KA, Bishopric NH. Modulation of cytokine-induced cardiac myocyte apoptosis by nitric oxide, Bak, and Bcl-x. Circ Res 1999;84:21e33. [33] Stenvinkel P, Rodriguez-Ayala E, Massy ZA, Qureshi AR, Barany P, Fellstrom B, et al. Statin treatment and diabetes affect myeloperoxidase activity in maintenance hemodialysis patients. Clin J Am Soc Nephrol 2006;1:281e7. [34] Kumar S, Raftery M, Yaqoob M, Fan SL. Anti-inflammatory effects of 3-hydroxy-3-methylglutaryl coenzyme a reductase inhibitors (statins) in peritoneal dialysis patients. Perit Dial Int 2007;27:283e7. [35] Mantuano E, Santi S, Filippi C, Manca-Rizza G, Paoletti S, Consani C, et al. Simvastatin and fluvastatin reduce interleukin6 and interleukin-8 lipopolysaccharide (LPS) stimulated production by isolated human monocytes from chronic kidney disease patients. Biomed Pharmacother 2007;61:360e5. [36] Paraskevas KI. Statin therapy in peritoneal dialysis patients: effects beyond lipid lowering. Int Urol Nephrol 2008;40: 165e70. [37] Wanner C, Krane V, Marz W, Olschewski M, Mann JF, Ruf G, et al. Atorvastatin in patients with type 2 diabetes mellitus undergoing hemodialysis. N Engl J Med 2005;353:238e48. [38] Fellstrom BC, Jardine AG, Schmieder RE, Holdaas H, Bannister K, Beutler J, et al. Rosuvastatin and cardiovascular events in patients undergoing hemodialysis. N Engl J Med 2009;360: 1395e407. [39] Mele ´ndez GC, McLarty JL, Levick SP, Du Y, Janicki JS, Brower GL. Interleukin 6 mediates myocardial fibrosis, concentric hypertrophy, and diastolic dysfunction in Rats. Hypertension 2010 Aug;56:225e31.