Is left ventricular hypertrophy a powerful predictor of progression to dialysis in chronic kidney disease?

Nephrol Dial Transplant (2010) 1 of 7 doi: 10.1093/ndt/gfq409 NDT Advance Access published July 13, 2010 Original Article Is left ventricular hypertrophy a powerful predictor of progression to dialysis in chronic kidney disease? Ernesto Paoletti 1, Diego Bellino 1, Anna Maria Gallina 2, Marco Amidone 1, Paolo Cassottana 3 and Giuseppe Cannella 1 1 Divisione di Nefrologia, Dialisi e Trapianto, 2 Dipartimento di Scienze della Salute dell Università and 3 Divisione di Cardiologia, Azienda Ospedaliera Universitaria San Martino, Genova, Italy Correspondence and offprint requests to: Ernesto Paoletti; E-mail: ernesto.paoletti@hsanmartino.it Abstract Background. The role of cardiovascular factors in predicting renal outcome has not been extensively elucidated. Herein, we report a prospective evaluation of the impact of left ventricular hypertrophy (LVH) on outcome in nondiabetic patients with chronic kidney disease (CKD). Methods. We studied 144 patients (99 men; age 62 ± 14 years) with stage 3 4 CKD, with baseline assessment of left ventricular mass index (LVMi) by echocardiography, estimated glomerular filtration rate (egfr) by MDRD equation, 24-h blood pressure profile and 24-h proteinuria. Combined end point was progression to ESRD requiring dialysis, or death within 5 years. Results. Forty-nine patients (34%) progressed to dialysis, 24 (17%) died, 57 (39%) were dialysis-free after 5 years and 14 were lost to follow-up. Multivariate Cox proportional hazards analysis showed that increased LVMi (HR 1.28, 95% CI 1.17 1.40 for each 10-g/m 2 increase, P < 0.0001) and reduced egfr (5% risk increase for each 1- ml/min reduction, P = 0.027) were the significant predictors of the combined end point in stage 3 CKD patients, whereas LVMi proved to be the only significant predictor of the combined end point in patients with stage 4 CKD (HR 1.19, 95% CI 1.09 1.31, P < 0.0001). The same analysis showed that LVMi was the only significant predictor of progression to dialysis in stage 3 CKD patients (HR 1.42, 95% CI 1.23 1.64 for each 10-g/m 2 increase, P < 0.0001), while a 20% increase in the risk of progression to ESRD was observed for each 10-g/m 2 increase in LVMi (P < 0.0001), and a 10% increase for each 1-mL/min reduction in egfr (P = 0.046) in patients with stage 4 CKD. When evaluating the predictive role of LVMi on outcome using AUC-ROC curves, the overall performance of the model including LVMi (AUC 0.877, 95% CI 0.8 0.954) was superior to the model including egfr (AUC 0.737, 95% CI 0.656 0.817) for the end point of progression to dialysis (P = 0.026, Hanley test). Conclusions. LVH proved to be the strongest predictor of the risk of progression to dialysis in non-diabetic CKD, especially among patients with less advanced renal dysfunction. Regardless of whether it is a simple marker or a pathogenetic factor, LVH encompasses all factors possibly affecting renal and general outcome in CKD patients. Keywords: chronic kidney disease; dialysis; left ventricular hypertrophy; mortality; outcome Introduction Epidemiological studies have reported a high prevalence of chronic kidney disease (CKD) in the general population since the KDOQI clinical practice guidelines were extensively adopted to identify and classify renal disease [1 3]. Only a minority of CKD patients, however, progress to end-stage renal disease (ESRD) requiring dialysis treatment. The main reason for this may be either the high incidence of CV death in CKD patients before they reach ESRD [4] or the heterogeneity in the risk of progressive renal function worsening due to complex interactions among genetic, environmental and pathogenetic factors [5]. Among these factors, reduced glomerular filtration rate and high urinary albumin excretion were identified as the main predictors of CKD progression to ESRD [6,7]. Furthermore, a recent study showed that combining the two data greatly improved the chances of identifying patients at increased risk of ESRD [8]. However, most studies lack information on many other potentially important risk factors, and especially on cardiovascular ones [9,10]. Establishing reliable and early indexes for identifying patients at high risk of developing ESRD should be a priority in the current effort to establish both clinical practice guidelines and plans for medical and financial management of CKD [5,11]. Therefore, the aim of this study was to evaluate the power of left ventricular hypertrophy (LVH) as compared with traditional risk factors in predicting the combined end point of progression to ESRD requiring renal replacement therapy (RRT), or death in a population of non-diabetic, The Author 2010. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org

2 E. Paoletti et al. stage 3 and 4 CKD patients who had previously been evaluated by echocardiography and 24-h ambulatory blood pressure monitoring (ABPM) for left ventricular mass determination and blood pressure profile assessment. Materials and methods Patients and study design Two hundred and forty-four patients from a cohort of 315 consecutive CKD patients who were referred to our renal unit for the first time were considered eligible for our previous study [12]. Exclusion criteria were diabetes in 35 patients, severe ischaemic heart disease or severe valvular disease in 23, inadequate acoustic window for echocardiographic examination in 5, malignancies in 2, and unwillingness to participate in the study in 6. All 244 patients had undergone both 24-h ABPM BP assessment and echocardiographic examination for left ventricular mass index (LVMi) measurement since the aim of our previous study was to assess the prevalence of LVH and its clinical correlate in a population of nondiabetic, CKD subjects [12]. From this historical cohort, we prospectively evaluated the 144 subjects with overt CKD (83 with stage 3 and 61 with stage 4 CKD) who had not yet reached ESRD at the time of the first observation. These patients underwent periodic clinical and biochemical profile evaluation which was scheduled on the basis of baseline CKD stage as well as on renal disease progression rate. The present study was designed with two primary end points, i.e. the combined end point of progression to ESRD requiring RRT or death from all cause, and progression to ESRD requiring RRT, all occurring within 5 years. The study protocol was drawn up according to the Declaration of Helsinki and in keeping with the recommendations issued by the ethical committee of our institution. Each patient gave informed consent. Measurements and definitions Descriptions of both the ABPM and the echocardiographic studies are reported elsewhere [12]. ABPM was performed using Spacelabs 90207 (SpaceLabs Inc., Redmond, WA, USA). We derived the average 24-h systolic blood pressure (SBP) and diastolic blood pressure (DBP), the awake SBP and DBP, and the asleep SBP and DBP from the ABPM data. Pulse pressure (PP) was calculated as SBP DBP for each set of BP measurements, either office or ABPM. Dipping was defined as a 10% reduction in night-time ambulatory BP compared with daytime ambulatory BP, whereas reverse dipping was defined as no reduction, or even an increase in night-time ABPM BP compared with daytime values [13]. Arterial hypertension was defined as an average BP 130/80 mmhg according to the criteria issued by the NKF-Kidney Disease Outcomes Quality Initiative Clinical Practice Guidelines on Hypertension and Antihypertensive Agents in CKD [14]. On the basis of ABPM records, patients were classified as normotensives, dippers, non-dippers or reverse dippers. Echocardiography was performed on each subject within 1 week either prior to or following ABPM using a two-dimensional guided M-mode echocardiograph according to the American Society of Echocardiography (ASE) recommendations [15]. Left ventricular mass (LVM) was calculated according to the formula, LVM = 0.80 1.04 [(IVS d +PW d + EDD) 3 EDD 3 ] + 0.6 g [16], and then indexed for body surface area (BSA) (LVMi). LVH was defined as LVMi >134 and >110 g/m 2 according to previously reported cut-off values for men and women, respectively [16]. Relative wall thickness (RWT) was calculated as the 2PW/EDD ratio, and a value of 0.44 was taken as the cut-off limit to define concentric (RWT >0.44) or eccentric (RWT 0.44) LVH [17]. The percent fractional shortening (FS) of the LV was calculated as EDD EDS / EDD 100. Systolic dysfunction was defined as an FS value 30%. Depending on whether LVH was present or not, as well as on the different LV geometry and FS value, patients were further divided into four groups, i.e. normal LV, concentric LVH, eccentric LVH and systolic dysfunction [17,18]. Serum creatinine, haemoglobin (Hgb) concentration, serum albumin, cholesterol, triglycerides, uric acid and immunoreactive parathyroid hormone (ipth) as well as 24-h urinary protein excretion rate were measured for each patient at least in triplicate and then averaged. Estimated glomerular filtration rate (egfr) was calculated according to the fourvariable MDRD equation [19]. Statistical analyses Analyses were performed using the statistical package SPSS 13.0 (SPSS, Chicago, IL, USA). A value of P < 0.05 was considered statistically significant. Data are presented as mean ± SD since all continuous variables were normally distributed. Normality of distribution was assessed by the Kolmogorov Smirnov test. The significance of association between each clinical and demographic variable (age, duration of either arterial hypertension or chronic kidney disease at the time of the first observation, number of antihypertensive medications, serum creatinine, Hgb, total cholesterol, triglycerides, serum albumin, ipth, daily urinary protein excretion, and each set of ABPM-derived BP values) was determined by calculating Spearman s correlation coefficient (r). Comparisons between groups were made by the unpaired t-test for continuous variables. Fisher s exact test or the chi-square test, when appropriate, was used for between-group comparisons of categorical variables. Cox proportional hazards analyses were performed to assess the hazard ratio (HR) and 95% CI for the risk of the combined end point of progression to ESRD requiring dialysis or death, and for death-censored risk of progression to dialysis. Both univariate and multivariate models were analysed separately for stage 3 CKD patients and for those with stage 4 CKD. The following independent variables were included in the regression model: age, pulse pressure, egfr, 24-h urinary protein excretion rate and LVMi. Kaplan Meier survival analysis was used to establish survival curves for combined end point-free survival and dialysis-free survival. To test the sensitivity and specificity of egfr and of LVMi in predicting both the combined end point and the end point of dialysis, we generated receiver-operating characteristic (ROC) curves, including area under the curve (AUC), and their 95% CI. Test performances for various models, including LVMi, egfr or both, were compared by the Hanley method. Results Of the 144 CKD subjects, 14 (10%) were lost to follow-up due to their moving to other cities or for unknown reasons, 49 (34%) progressed to ESRD requiring RRT within the 5-year observation period, 24 (17%) died, and 57 (39%) were alive and RRT-free at the end of the 5-year follow-up period. Table 1 shows the baseline demographic and clinical data of the study population. No significant differences were found between the 130 patients who were followed up for 5 years and the 14 who were lost to follow-up. At baseline, 24-h PP was directly associated with age (r = 0.56; P < 0.0001), 24-h proteinuria (r = 0.19; P = 0.026) and LVMi (r = 0.18; P = 0.022), whereas both 24-h proteinuria and LVMi were inversely associated with egfr (P = 0.039 and P = 0.045, respectively). Furthermore, a significant inverse association was found between LVMi and FS (P = 0.004). Table 2 shows a comparison analysis of demographic and clinical data between the 73 subjects who reached the combined end point within the 5-year observation period and the 57 subjects who were still alive and dialysis-free at the end of this period. Subjects who were alive at end point had higher egfr and lower baseline LVMi than patients who either had started dialysis or had died in the meantime (both P < 0.0001, unpaired t-test). Moreover, a higher proportion of subjects reaching the end point of dialysis or death had stage 4 CKD and greater prevalence of LVH as well as a more unfavourable LV geometric pattern such as eccentric

LVH and CKD progression 3 Table 1. Baseline demographic and clinical parameters of the study population Age, years (range) 62 (17 76) Gender, M/F (%) 99 (69)/45 (31) Diagnosis of renal disease, GN/TI/HPT/APKD/other (%) 54 (37)/11 (8)/62 (43)/9 (6)/8 (6) egfr, ml/min 32.9 ± 12.3 Stage CKD, 3/4 (%) 83 (58)/61 (42) Hgb, g/dl 11.9 ± 1.8 Serum albumin, g/dl 3.9 ± 0.7 Total cholesterol, mg/dl 229 ± 59 Triglycerides, mg/dl 206 ± 150 Uric acid, mg/dl 7.0 ± 2.3 ipth, pg/ml 82 ± 64 24-h SBP, mmhg 139 ± 16 24-h DBP, mmhg 81 ± 9 24-h PP, mmhg 58 ± 14 Dippers/non-dippers/reverse dippers (%) 11 (8)/88 (61)/45 (31) Patients on RAAS blockers/other antihypertensives/statins (%) 46 (32)/101 (70)/68 (47) 24-h urinary protein excretion rate, g/day 1.1 ± 1.78 LVMi, g/m 2 BSA 145 ± 41 FS, % 37.9 ± 8.4 LV geometry, normal/concentric LVH/eccentric LVH/systolic dysfunction (%) 52 (36)/37 (26)/38 (26)/17 (12) Comparison between groups was made by unpaired t-test for continuous variables. Fisher s exact test or the chi-square test, when appropriate, was used for between-group comparison of categorical variables. GN, glomerulonephritis; TI, chronic tubulointerstitial nephropathy; HPT, hypertensive nephropathy; APKD, polycystic kidney disease; egfr, estimated glomerular filtration rate; CKD, chronic kidney disease; Hgb, haemoglobin; ipth, intact parathyroid hormone; SBP, systolic blood pressure; DBP, diastolic blood pressure; PP, pulse pressure; LVMi, left ventricular mass index; BSA, body surface area; LVH, left ventricular hypertrophy; FS, left ventricular fractional shortening; LVH, left ventricular hypertrophy. LVH or systolic dysfunction compared with surviving dialysis-free subjects (all P < 0.0001, Fisher s exact test or chisquare test). In contrast, no significant differences were observed in the two groups regarding age, gender, and the prevalence of arterial hypertension, dipping habit, albuminuria, serum albumin and anaemia. Lastly, higher LVMi (167 ± 41 g/m 2 ) and greater prevalence of stage 4 CKD (65%) were the only significant differences we observed in the 49 patients who reached ESRD and who started dialysis as compared with the 57 subjects who were still alive, dialysis-free after the 5-year observation period (both P < 0.0001; unpaired t-test and Fisher s exact test, respectively). Risk of progression to dialysis or death (combined end point) Univariate Cox regression analysis showed a 27% increase in the risk of the combined end point for each 10-g/m 2 in- Table 2. Comparison between baseline demographic and clinical parameters of 57 CKD patients alive and dialysis-free after 5 years, and 73 CKD patients reaching the combined end point within the 5-year observation period Alive, dialysis-free Combined end point P n 57 73 Age, years 61 ± 17 63 ± 11 0.32 Gender, M/F 36/21 53/20 0.26 Diagnosis of renal disease, GN/TI/HPT/APKD/other 18/6/27/3/3 29/5/30/6/3 0.75 egfr, ml/min 39.4 ± 11.5 28.1 ± 10.8 <0.0001 Stage CKD, 3/4 45/12 31/42 <0.0001 Hgb, g/dl 12.3 ± 2.0 11.7 ± 1.6 0.08 Serum albumin, g/dl 3.9 ± 0.8 4.0 ± 0.7 0.45 Total cholesterol, mg/dl 232 ± 68 226 ± 54 0.54 Triglycerides, mg/dl 199 ± 139 210 ± 162 0.68 Uric acid, mg/dl 7.1 ± 2.4 6.9 ± 2.1 0.61 ipth, pg/ml 90 ± 50 76 ± 79 0.22 24-h SBP, mmhg 138 ± 13 140 ± 18 0.43 24-h DBP, mmhg 82 ± 11 81 ± 9 0.49 24-h PP, mmhg 56 ± 13 60 ± 15 0.17 Dippers/non-dippers/reverse dippers 5/35/17 3/43/27 0.44 Patients on RAAS blockers/other antihypertensives/statins 18/41/26 19/49/32 0.9 24-h urinary protein excretion rate, g/day 1.05 ± 2.0 1.04 ± 1.57 0.9 LVMi, g/m 2 BSA 120 ± 30 165 ± 40 <0.0001 LVH prevalence, yes/no 18/39 65/8 <0.0001 FS, % 39.5 ± 7.1 36.0 ± 8.8 0.06 LV geometry, normal/concentric LVH/eccentric LVH/systolic dysfunction 39/8/8/2 8/26/25/14 <0.0001

4 E. Paoletti et al. Table 3. Cox proportional hazard analysis of the combined end point progression to dialysis or death in 83 stage 3 CKD patients Age, 1-year increase 1.014 (0.986 1.043) 0.32 Proteinuria, 1-g/24 h increase 1.069 (0.825 1.384) 0.61 24-h PP, 1-mmHg increase 1.017 (0.992 1.043) 0.17 egfr, 1-mL/min increase 0.94 (0.898 0.985) 0.01 LVMi, 10-g/m 2 increase 1.268 (1.166 1.38) <0.0001 Age, 1-year increase 1.019 (0.989 1.05) 0.21 Proteinuria, 1-g/24 h increase 1.27 (0.937 1.723) 0.12 24-h PP, 1-mmHg increase 1.018 (0.952 1.048) 0.26 egfr, 1-mL/min increase 0.948 (0.904 0.994) 0.027 LVMi, 10-g/m 2 increase 1.278 (1.167 1.399) <0.0001 crease in baseline LVMi and a 6% increase for each 1-mL reduction in baseline egfr for patients with stage 3 CKD (Table 3), and an 18% increase for each 10-g/m 2 increase in baseline LVMi and a 9% increase for each 1-mL reduction in egfr in patients with stage 4 CKD (Table 4). showed that LVMi (HR 1.28; 95% CI 1.17 1.4; P < 0.0001) and, to a lesser extent, egfr (HR 0.95; 95% CI 0.90 0.99; P = 0.027) were significant predictors of the risk of combined end point in stage 3 CKD patients (data adjusted for age, pulse pressure and proteinuria) (Table 3), whereas LVMi (HR 1.194; 95% CI 1.09 1.31) proved to be the only significant predictor of the combined end point in patients with stage 4 CKD (data adjusted for age, pulse pressure, egfr and proteinuria) (Table 4). Figure 1 shows Kaplan Meier curves for combined end point-free survival in the 144 stage 3 4 CKD patients subdivided according to the presence or absence of LVH at baseline (Figure 1). When evaluating the role of the various baseline LV geometric patterns in predicting the combined end pointfree survival, patients with systolic dysfunction showed the worst survival rate according to Kaplan Meier analysis (P < 0.001, log-rank test). Table 4. Cox proportional hazard analysis of the combined end point progression to dialysis or death in 61 stage 4 CKD patients Age, 1-year increase 1.004 (0.984 1.023) 0.71 Proteinuria, 1-g/24 h increase 0.957 (0.82 1.113) 0.57 24-h PP, 1-mmHg increase 1.004 (0.983 1.025) 0.72 egfr, 1-mL/min increase 0.909 (0.838 0.987) 0.023 LVMi, 10-g/m 2 increase 1.177 (1.09 1.27) <0.0001 Age, 1-year increase 1.001 (0.97 1.029) 0.9 Proteinuria, 1-g/24 h increase 1.116 (0.917 1.357) 0.27 24-h PP, 1-mmHg increase 1.012 (0.958 1.02) 0.45 egfr, 1-mL/min increase 0.945 (0.87 1.028) 0.19 LVMi, 10-g/m 2 increase 1.194 (1.09 1.31) <0.0001 Fig. 1. Kaplan Meier analysis of combined end point-free survival in patients with LVH and in those with normal LVMi at baseline (log-rank P < 0.001). Risk of progression to dialysis Univariate Cox proportional hazards analysis showed that each 10-g/m 2 increase in baseline LVMi was associated with a 28% increase in the risk of progression to ESRD requiring dialysis within the 5-year follow-up period in patients with stage 3 CKD (Table 5), whereas a 19% increase in the HR for each 10-g/m 2 increase in baseline LVMi and a 13% increase for each 1-mL reduction in egfr were observed in patients with stage 4 CKD (Table 6). proved that increased LVMi was the only significant predictor of the risk of progression to dialysis in patients with stage 3 CKD (data adjusted for age, egfr, pulse pressure and 24-h urinary protein excretion rate) (Table 5), whereas a 20% increase in the risk for each 10-g/m 2 increase in baseline LVMi and a 10% increase for each 1-mL reduction in egfr were observed in patients with stage 4 CKD (data adjusted for age, pulse pressure and proteinuria) (Table 6). Figure 2 shows Kaplan Meier curves for dialysis-free survival in the 144 CKD patients with and without LVH at baseline (Figure 2). Table 5. Cox proportional hazard analysis of progression to dialysis in 83 stage 3 CKD patients Age, 1-year increase 1.005 (0.971 1.041) 0.77 Proteinuria, 1-g/24 h increase 1.063 (0.75 1.507) 0.73 24-h PP, 1-mmHg increase 1.025 (0.935 1.063) 0.22 egfr, 1-mL/min increase 0.961 (0.905 1.02) 0.19 LVMi, 10-g/m 2 increase 1.285 (1.147 1.438) <0.0001 Age, 1-year increase 1.027 (0.989 1.067) 0.16 Proteinuria, 1-g/24 h increase 1.292 (0.848 1.969) 0.23 24-hr PP, 1-mmHg increase 1.061 (0.966 1.167) 0.22 egfr, 1-mL/min increase 0.96 (0.90 1.02) 0.26 LVMi, 10-g/m 2 increase 1.42 (1.23 1.64) <0.0001

LVH and CKD progression 5 Table 6. Cox proportional hazard analysis of progression to dialysis in 61 stage 4 CKD patients Age, 1-year increase 1.007 (0.973 1.013) 0.48 Proteinuria, 1-g/24 h increase 0.925 (0.77 1.113) 0.41 24-h PP, 1-mmHg increase 1.003 (0.973 1.021) 0.78 egfr, 1-mL/min increase 0.867 (0.783 0.962) 0.006 LVMi, 10-g/m 2 increase 1.194 (1.093 1.304) <0.0001 Age, 1-year increase 1.02 (0.948 1.052) 0.23 Proteinuria, 1-g/24 h increase 1.07 (0.835 1.37) 0.6 24-h PP, 1-mmHg increase 1.009 (0.954 1.03) 0.66 egfr, 1-mL/min increase 0.902 (0.812 1.008) 0.046 LVMi, 10-g/m 2 increase 1.196 (1.073 1.33) 0.001 The role of LVMi and egfr in predicting outcome AUC-ROC curves showed that a model, which included both LVMi and egfr, was highly predictive of both the combined end point (AUC 0.854, 95% CI 0.772 0.935) and the progression to dialysis (AUC 0.884, 95% CI 0.814 0.955) (Figure 3). Moreover, a comparison of the AUC-ROC curves by the Hanley test showed that the model including both LVMi and egfr was more predictive than models including either LVMi or egfr alone in predicting both the combined end point (P = 0.004 vs LVMi alone and P = 0.001 vs egfr alone) and the progression to dialysis (P < 0.001 vs LVMi and P = 0.019 vs egfr). Lastly, the model including LVMi provided significantly higher sensitivity and specificity than the model with egfr in predicting progression to dialysis (P = 0.026). Discussion Fig. 2. Kaplan Meier analysis of dialysis-free survival in patients with LVH and in those with normal LVMi at baseline (log-rank P < 0.001). Fig. 3. The AUC-ROC curves for combined end point (A) and the end point of progression to dialysis (B) as predicted by a model including LVMi alone, egfr alone or both. Clinical predictors in both models included age, 24-h proteinuria and 24-h pulse pressure. Prospective analysis of our historical cohort of non-diabetic CKD patients [12] showed that baseline LVH was the strongest predictor of a composite end point consisting of either ESRD requiring dialysis, or death. LVH has indeed been independently associated with adverse outcome in the general population and in renal patients as well [20 22]. The most relevant and novel finding of our study, however, is that LVH also proved to be the most significant predictor of ESRD requiring RRT, despite the inclusion of egfr values in the Cox proportional hazards model, as well as the only predictor of the progression to dialysis in patients with less advanced CKD. Although we do not present a definite explanation for this finding, we do hypothesize that LVH could be a marker of the risk of renal disease progression towards the end stage because the same pathogenetic factors that are involved in kidney damage are also responsible for the increase in both LV wall thickness and internal dimension. In our cohort, baseline LVH was in fact significantly associated with elevated PP and older age, whereas no associations were shown with egfr, the degree of anaemia, dyslipidaemia, ipth or serum albumin [12]. However, the prevalence of arterial hypertension in our cohort was similar in patients reaching ESRD and in dialysis-free survivors. Moreover, 24-h BP values did not significantly predict the risk of developing ESRD and dialysis, thus ruling out the possibility that inadequate BP control may have played a major role in both cardiac hypertrophy and CKD progression in our patients. Furthermore, cardiac and renal involvement could be the effect of endothelial dysfunction inducing generalized vascular damage in CKD patients, which is consistent with findings from the cohort enrolled in the LIFE Study which demonstrated a significant association of LVH with urinary albumin excre-

6 E. Paoletti et al. tion [23]. Accordingly, a significant association was indeed found in our sample between increased PP, an indirect expression of vascular stiffness, and both LVMi and proteinuria. Recent reports that are consistent with this pathogenetic hypothesis have shown that NT-proBNP and proanp, which are recognized markers of cardiac function impairment, proved to be strong predictors of CKD progression to ESRD in non-diabetic patients [24,25]. LVH might also exert a direct pathogenetic effect on the progression of CKD to ESRD. Reports exist showing that LVH may be associated with subclinical worsening of systolic LV function [26 28]. Moreover, mild impairment in systolic LV function reportedly predicts the risk of adverse CV outcome in dialysis patients [29]. Interestingly, in our cohort, a significant, inverse relationship was shown between LVMi and left ventricular fractional shortening. Moreover, consistent with previous reports in dialysed patients [18], our patients with LV systolic dysfunction showed worse outcome. We observed a direct association between increased PP and LVH, thus confirming the role of arterial stiffness and the consequent increase in velocity of backward reflected waves to the LV in the pathogenesis of cardiac hypertrophy [30]. We cannot rule out that a mild reduction in renal perfusion induced by slightly impaired LV systolic function associated with pathological, highly pulsatile perfusion in the kidney microvasculature as an effect of reduced elasticity of conduit arteries [31] may have played a role in promoting a progressive, further reduction in renal function of patients with pre-existing CKD. Another finding of this study is that increased urinary protein excretion rate did not prove to be a significant predictor of the risk of progression to dialysis in our cohort. We speculate that this might be the consequence of the stronger pathogenetic role played by LVH which possibly masked the effect of proteinuria in predicting adverse renal outcome. Lastly, no predictive role for anaemia was found in our sample. This is consistent with an interventional study that showed neither regression of LVH nor slowing of renal disease progression in CKD patients in whom complete anaemia correction was achieved by epoetin therapy [32]. Of course, some limitations must be acknowledged in our study. First of all, sample size is small, and 10 patients were missed and lost to follow-up. Secondly, there is great heterogeneity in patients age and higher prevalence of men, two factors which may affect LVH in CKD. Furthermore, we have no data concerning the behaviour of LVMi during the observation period, and we did not evaluate the potential role of changes in some predictors and the effect of intervention on such factors during follow-up in counteracting both the progression to ESRD and the risk of death. Lastly, we did not evaluate the role of some potential predictors of adverse outcome in CKD, such as NT-proBNP which reportedly predicts LVH [33], and high-sensitivity C-reactive protein whose association with LVH in CKD however remains controversial [34]. The aim of our study was not to investigate the pathogenesis of LVH in CKD, but its main objective was to assess the impact of cardiac hypertrophy on both CV and renal outcome in CKD patients as compared with well-known risk factors of progression to ESRD and death, namely egfr, proteinuria and arterial hypertension. This is indeed the first long-term observation in CKD patients to point out the pivotal role of LVH in predicting not only the risk of death but also of ESRD requiring dialysis. On the basis of our findings, it is conceivable that LVH agglomerates and encompasses all the potential mechanisms involved in inducing either ESRD or death in patients with overt CKD. Accordingly, systematic assessment of morphology and function of the LV should be considered a key tool in order to obtain prognostic information not only about the CV outcome of CKD patients but also about the progression of their renal disease. This could be especially important in less advanced renal disease patients for whom intervention might still be effective in slowing down progression to the end stage. Conflict of interest statement. None declared. References 1. Hallan SI, Coresh J, Astor BC et al. International comparison of the relationship of chronic kidney disease prevalence and ESRD risk. J Am Soc Nephrol 2006; 17: 2275 2284 2. Coresh J, Selvin E, Stevens LA et al. Prevalence of chronic kidney disease in the United States. JAMA 2007; 298: 2038 2047 3. Stevens PE, O Donoghue DJ, de Lusignan S et al. Chronic kidney disease management in the United Kingdom: NEOERICA project results. Kidney Int 2007; 72: 92 99 4. Keith DS, Nichols GA, Gullion CM et al. Longitudinal follow-up and outcomes among a population with chronic kidney disease in a large managed care organization. Arch Intern Med 2004; 164: 659 663 5. Taal MW, Brenner BM. Renal risk scores: progress and prospects. Kidney Int 2008; 73: 1216 1219 6. Ishani A, Grandits GA, Grimm RH et al. Association of single measurements of dipstick proteinuria, estimated glomerular filtration rate, and hematocrit with 25-year incidence of end-stage renal disease in the multiple risk factor interventional trial. J Am Soc Nephrol 2006; 17: 1444 1452 7. van der Velde M, Halbesma N, de Charro FT et al. Screening for albuminuria identifies individuals at increased renal risk. J Am Soc Nephrol 2009; 20: 852 862 8. Hallan SI, Ritz E, Lydersen S et al. Combining GFR and albuminuria to classify CKD improves prediction of ESRD. JAmSocNephrol 2009; 20: 1069 1077 9. Iseki K, Kinjo K, Iseki C et al. Relationship between predicted creatinine clearance and proteinuria and the risk of developing ESRD in Okinawa, Japan. Am J Kidney Dis 2004; 44: 806 814 10. Lea J, Greene T, Hebert L et al. The relationship between magnitude of proteinuria reduction and the risk of end-stage renal disease: results of the African American Study of Kidney Disease and Hypertension. Arch Intern Med 2005; 165: 947 953 11. Crowe E, Halpin D, Stevens P. Guideline Development Group. Early identification and management of chronic kidney disease: summary of NICE guidance. BMJ 2008; 337: a153 12. Paoletti E, Bellino D, Cassottana P et al. Left ventricular hypertrophy in nondiabetic predialysis CKD. Am J Kidney Dis 2005; 46: 320 327 13. O Brien E, Sheridan J, O Malley K. Dippers and non-dippers. Lancet 1988; 2: 397 14. National Kidney Foundation: K/DOQI Clinical Practice Guidelines on Hypertension and Antihypertensive agents in Chronic Kidney Disease. Am J Kidney Dis 2004; 43: S1 S290 15. Sahn DJ, De Maria A, Kisslo J et al. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 1978; 58: 1072 1083 16. Devereux RB, Alonso DR, Lutas EM et al. Echocardiographic assessment of left ventricular hypertrophy: comparison to necroscopy findings. Am J Cardiol 1986; 57: 450 458

LVH and CKD progression 7 17. Ganau A, Devereux RB, Roman MJ et al. Patterns of left ventricular hypertrophy and geometric remodeling in essential hypertension. J Am Coll Cardiol 1992; 19: 1550 1558 18. Parfrey PS, Foley RN, Harnett JD et al. Outcome and risk factors for left ventricular disorders in chronic uremia. Nephrol Dial Transplant 1996; 11: 1277 1285 19. Levey AS, Greene T, Kusek JWet al. A simplified equation to predict glomerular filtration rate from serum creatinine (abstract). J Am Soc Nephrol 2000; 11: A0828 20. Levy D, Garrison RJ, Savage DD et al. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med 1990; 322: 1561 1566 21. London GM, Pannier B, Guerin AP et al. Alterations of left ventricular hypertrophy in and survival of patients receiving hemodialysis: follow-up of an interventional study. J Am Soc Nephrol 2001; 12: 2759 2767 22. Paoletti E, Cassottana P, Bellino D et al. Left ventricular geometry and adverse cardiovascular events in chronic hemodialysis patients on prolonged therapy with ACE inhibitors. Am J Kidney Dis 2002; 40: 728 736 23. Olsen MH, Wachtell K, Borch-Johnsen K et al. A blood pressure independent association between glomerular albumin leakage and electrocardiographic left ventricular hypertrophy. The LIFE Study. Losartan Intervention for Endpoint Reduction. JHumHypertens 2002; 16: 591 595 24. Spanaus KS, Kronenberg S, Ritz E et al. B-type natriuretic peptide concentrations predict the progression of nondiabetic chronic kidney disease: the Mild-to-Moderate Kidney Disease Study. Clin Chem 2007; 53: 1264 1272 25. Dieplinger B, Mueller T, Kollerits B et al. Pro-A-type natriuretic peptide and pro-adrenomedullin predict progression of chronic kidney disease: the MMKD Study. Kidney Int 2009; 75: 408 414 26. de Simone G, Kitzman DW, Palmieri V et al. Association of inappropriate left ventricular mass with systolic and diastolic dysfunction. The HyperGEN Study. Am J Hypertens 2004; 17: 828 833 27. Drazner MH, Rame JE, Marino EK et al. Increased left ventricular mass is a risk factor for the development of a depressed left ventricular ejection fraction within five years. The Cardiovascular Health Study. J Am Coll Cardiol 2004; 43: 2207 2215 28. Schrier RW. Role of diminished renal function in cardiovascular mortality. J Am Coll Cardiol 2006; 47: 1 8 29. Zoccali C, Benedetto FA, Tripepi G et al. Left ventricular systolic function monitoring in asymptomatic dialysis patients: a prospective cohort study. J Am Soc Nephrol 2006; 17: 1460 1465 30. Marchais SJ, Guerin AP, Pannier BM et al. Wave reflections and cardiac hypertrophy in chronic uremia: influence of body size. Hypertension 1993; 22: 876 883 31. O Rourke MF, Safar ME. Relationship between aortic stiffening and microvascular disease in brain and kidney. Cause and logic of therapy. Hypertension 2005; 46: 200 204 32. Roger SD, McMahon LP, Clarkson A et al. Effects of early and late intervention with epoetin alpha on left ventricular mass among patients with chronic kidney disease (stage 3 or 4): results of a randomized clinical trial. J Am Soc Nephrol 2004; 15: 148 156 33. Khan IA, Fink J, Nass C et al. N-terminal pro-b-type natriuretic peptide and B-type natriuretic peptide for identifying coronary artery disease and left ventricular hypertrophy in ambulatory chronic kidney disease patients. Am J Cardiol 2006; 97: 1530 1534 34. Wang AY, Lam CW, Wang M et al. Diagnostic potential of serum biomarkers for left ventricular abnormalities in chronic peritoneal dialysis patients. Nephrol Dial Transplant 2009; 24: 1962 1969 Received for publication: 1.12.09; Accepted in revised form: 18.6.10