Patients with mild to moderate impairment in renal function are at increased risk for adverse cardiovascular (CV) outcomes. After myocardial infarction (MI), any renal dysfunction is associated with increased risk of CV death, heart failure (HF), sudden death, stroke, or recurrent MI (1- 6). Cardiovascular mortality is 10 to 30 times higher in patients with end-stage renal disease (ESRD) as compared with the general population and approaches 50% at 2 years after MI (1). End-stage renal disease is associated with a variety of cardiac alterations including left ventricular hypertrophy (LVH), LV dilation, and reduction in systolic and diastolic function, with only 16% of new dialysis patients presenting with normal cardiac morphology and function (7- 8).

Although patients with reduced renal function are at significantly increased risk of all CV events, including death, recurrent MI, and HF after MI, the reasons for this increased risk have been less clear. Increased vascular events clearly play an important role, but the increased risk of HF cannot be explained by vascular events alone. This raises the possibility that alterations in cardiac function might play a central role in increased risk of HF development and CV mortality.

To assess whether alterations in cardiac structure or function contribute to the increased risk associated with reduced renal function in the post-MI patient, we studied patients enrolled in the echocardiographic substudy of the VALIANT (VALsartan In Acute myocardial iNfarcTion) study.

The VALIANT study was designed to test the hypothesis that the angiotensin receptor blocker valsartan, either alone or in combination with the proven angiotensin-converting enzyme inhibitor captopril, would be superior or not inferior to a proven dose of captopril in reducing CV morbidity or mortality after MI (9). Patients with an acute MI complicated by HF, LV systolic dysfunction (ejection fraction [EF] <35% on echocardiography or <40% on contrast angiography), or both were randomly assigned between 12 h and 10 days to treatment with either captopril (target dose: 50 mg 3 times daily), valsartan (target dose: 160 mg twice daily), or their combination (target doses: captopril 50 mg 3 times daily and valsartan 80 mg twice daily) in a 1:1:1 ratio (9). Patients with a baseline serum creatinine level of at least 2.5 mg/dl (221 μmol/l) were excluded (9). The median duration of follow-up was 24.7 months. A total of 610 patients were enrolled in the echocardiographic substudy. Patients underwent 2-dimensional echocardiography at a mean time of 5.0 ± 2.5 days after the index MI (baseline) and at 1 and 20 months (10). Details of patient characteristics have been previously described. Inclusion and exclusion criteria were identical to those of the main VALIANT trial, and the demographics of the echocardiographic participants were similar to the overall study group but differed from the main cohort minimally with respect to age and baseline blood pressure (10).

All echocardiographic studies were analyzed in a core laboratory. Before analysis, 7 patients were excluded because of insufficient echocardiographic images. Echocardiograms were performed at 5 days (baseline) (n = 603), 1 month (n = 544), and 20 months after MI (n = 428). A total of 59 patients in the initial cohort did not have 1-month echocardiograms, including 21 patients who died before the 1-month echocardiogram. The 20-month echocardiogram was not performed in 175 patients present in the initial cohort, including 98 patients who died before the final visit.

Echocardiograms from videotape were digitized and analyses were performed on an offline analysis workstation. The LV endocardial borders were manually traced by a single experienced observer at end-diastole and -systole at the mitral and papillary short-axis level and apical 4- and 2-chamber views from 3 separate cardiac cycles. The LV volumes were derived according to the modified biplane Simpson’s rule in the apical 4- and 2-chamber views, and EF was calculated in the standard fashion from LV end-diastolic volume (EDV) and LV end-systolic volume (ESV). Infarct segment length was assessed by manually tracing the akinetic or dyskinetic segment. Right ventricular (RV) function, expressed as the RV fractional area change (RVFAC), was assessed quantitatively as the percent change in cavity area from end-diastole to end-systole. Baseline LV mass was calculated by the American Society of Echocardiography (ASE) recommended formula for estimation of LV mass from LV linear dimensions and indexed to body surface area (11). The LVH was considered present when echocardiographically derived baseline left ventricular mass index (LVMI) was >115 g/m² for men and >95 g/m² for women (11).

Mitral regurgitation (MR) was categorized by mapping and tracing the jet area (obtained with color Doppler imaging) occupying the left atrium in 4- and 2-chamber views at end-diastole and -systole from 3 separate cardiac cycles and was expressed as a proportion of left atrial (LA) area. The MR was considered mild when regurgitant jet area occupied >5% and <20% of the LA area, moderate when >20% and <40%, and severe when >40% (12- 13). The presence of an eccentric jet raised the grade of MR by 1 degree (14). The LA volume was assessed by the biplane area-length method from apical 2- and 4-chamber views at end-systole from the frame preceding mitral valve opening. The left atrial volume index (LAVI) was calculated as LA volume/body surface area (ml/m²). Mitral flow velocity was assessed by pulsed wave Doppler study from the apical 4-chamber view, by positioning the sample volume at the tip of the mitral leaflets.

Echocardiographic measurements were made in triplicate by a single experienced observer blinded to outcome data and treatment assignment with quantitative analysis software. Reproducibility was assessed after studies were randomly chosen and reanalyzed with the observer blinded to the initial results. The mean signed differences and the SD of the differences between measurements were calculated (Bland-Altman method), and the variability was expressed as the SD of the difference divided by the mean measurement. The coefficient of variability based on the reproducibility assessment was 8.3%, 2.7%, 3.0%, 5.3%, and 1% for LV volume, LV mass, LA volume, RVFAC, and MR assessment, respectively.

Patients were categorized according to the eGFR at baseline with the use of the 4-component Modification of Diet in Renal Disease (MDRD) study equation incorporating age, race, gender, and serum creatinine level (6,15):$eGFR = 186×(serum creatininelevel[in mg/dl])−1.154$$×$$(age [in years])−0.203$

For women and blacks, the product of this equation was multiplied by a correction factor of 0.742 and 1.21, respectively. The distribution of the eGFR was divided into 4 groups (<45.0, 45.0 to 59.9, 60.0 to 74.9, and ≥75.0 ml/min/1.73 m²) (6,16). The present stratification of the baseline eGFR is based on the previous report of Anavekar et al. (6) from the main VALIANT cohort.

Continuous data were expressed as mean ± SD. Among the 4 categories of eGFR, categorical variables were analyzed with the chi-square test. Continuous variables were compared with the nonparametric Wilcoxon rank sum test for trend across ordered groups. In addition, changes in echocardiographic parameters between baseline and 20 months were also compared across the 4 ordered groups with the nonparametric Wilcoxon rank sum test for trend. Multivariate linear regression analysis was performed to investigate the relationship between eGFR and the echocardiographic parameters adjusting for age, history of hypertension, and history of diabetes. Event rates were calculated with survival time data and expressed as events/100 patient-years. To determine the independent value of baseline eGFR, we used a multivariable Cox proportional hazards model. The primary end point for the main VALIANT study was death from any cause (9). Secondary end points included death from CV causes, congestive HF, recurrent MI, resuscitation after cardiac arrest, stroke, and a CV composite of these. Clinical outcomes including hospital stay for HF were adjudicated by an independent Clinical Endpoints Committee. Additionally, for the purpose of this analysis, clinical outcomes included a composite of death from any cause and hospital stay for HF, which was defined as the unplanned treatment of new or worsening HF requiring the use of intravenous diuretic, inotrope, or vasodilator drugs during any hospital admission or overnight stay in a healthcare facility.

The adjustment model included predictors of mortality identified from the overall VALIANT study: age (in years), gender, primary percutaneous transluminal coronary angioplasty (PTCA) after MI, atrial fibrillation complicating MI, history of diabetes, history of hypertension, prior MI, Killip class, and a history of chronic obstructive pulmonary disease (COPD). In addition to the aforementioned risk factors, adjustment was also made for the echocardiographic variable of interest with each variable entering separately into the model and treatment assignment. The echocardiographic variables were: LV ESV, LVEF, LAVI, ratio of MR area to LA area, RVFAC, and LVMI. All p values were 2-sided; p < 0.05 was used to determine statistical significance. Statistical analyses were performed with STATA software, version 8.2 (Stata Corp., College Station, Texas).

Baseline eGFR for the 603 patients in the VALIANT echocardiographic cohort was normally distributed, and the mean ± SD eGFR was 71.4 ± 20.5 ml/min/1.73 m², which was similar to the nonechocardiographic cohort (eGFR 70.2 ± 21.3). Despite the use of a serum creatinine level of at least 2.5 mg/dl as an exclusion criterion, 180 (29.9%) patients met eGFR criteria for chronic kidney disease (eGFR <60 ml/min/1.73 m²) (16).

Reduced eGFR was associated with higher rates of prior MI, prior hypertension, prior diabetes mellitus, and prior congestive HF (Table 1). Patients in the lower eGFR categories were less likely to be treated with beta-blocker drugs and less likely to have received coronary revascularization. Although lower eGFR was associated with increasing age and female gender, these variables were used in the determination of eGFR.

The LVEF, RVFAC, and infarct segment lengths were similar across the 4 categories of eGFR (Table 2). Decreasing eGFR was associated with a significantly lower EDV, a higher baseline LAVI ((Figure 1)A and Figure 1B), and a higher LVMI (Table 2). The LVH increased in a stepwise fashion with decreasing eGFR (p for trend <0.001), as did the LV mass/EDV ratio (p for trend <0.001) (Figure 1C). A trend toward a lower ESV was also observed with decreasing eGFR but did not achieve statistical significance. The LVMI (p = 0.002), LV mass/EDV ratio (p = 0.016), LAVI (p = 0.024), and MR (p = 0.019) remained significantly associated with reduced eGFR after adjusting for age, history of diabetes, and hypertension.

Quantitative assessment of MR by color Doppler was available in 496 patients. Lower eGFR was associated with a higher proportion of patients presenting with MR (p = 0.002) after MI and with a higher ratio of MR jet area to LA area (Table 2). There were no differences in transmitral flow and mitral deceleration time at baseline across the 4 groups (Table 2).

Paired echocardiographic studies between baseline and 20 months were available in 428 patients. There were no significant differences in the magnitude of the change in LV EDV, LV ESV, and LVEF between patients in either category of eGFR from baseline to 20 months with the exception of LAVI, which significantly increased in patients with reduced eGFR (Table 3).

In this cohort, 103 (17.1%) patients died (93 CV), 63 (10.4%) patients experienced recurrent MI, 17 (2.8%) patients experienced stroke, 95 (15.7%) patients were hospitalized for HF, and 162 (27%) patients either died or developed HF. To assess whether abnormalities of cardiac structure and function could account for the influence of reduced eGFR on death or development of HF, we created a multivariable Cox proportional hazards model including renal function, measures of cardiac structure and function (ESV, EF, LAVI, ratio of MR jet area to LA area), and known clinical covariates (age, history of MI, gender, history of hypertension, Killip class, history of COPD, history of diabetes, atrial fibrillation after MI, and primary PTCA for index MI). Worsening renal function was associated with increased risk of death or HF, even after adjusting for measures of cardiac function, clinical covariates, and treatment assignment (Table 4). However, the addition of LVMI to the model attenuated the association between worsening renal function and death or hospital stay for HF (Table 3). Both LVMI (HR 1.02/each 1 g/m² increase in LVMI; 95% confidence interval [CI] 1.02 to 1.03, overall model chi-square 176.1) and LAVI (HR 1.04/each 1 ml/m² increase in LAVI; 95% CI 1.02 to 1.06, overall model chi-square 110.0) conferred additional prognostic information in a multivariable model (Table 4).

In this analysis, we assessed the relationship between renal function and cardiac structure and function in high-risk survivors of MI. The increased risk of CV mortality and hospital stay for HF in patients with reduced renal function could not be explained by reduced systolic function, because left (and right) ventricular function was similar across eGFR categories. Indeed, we observed that patients with worse renal function had smaller, more hypertrophied hearts. Lower eGFR was associated with a lower LV EDV and ESV and higher LV mass, LV mass/EDV ratio, and LA volume (Figure 1). A greater proportion of the patients in the lower eGFR category had LVH.

A variety of morphological changes have been described in patients with chronic kidney disease and in patients with renal failure on renal replacement therapy. These ultrastructure changes, sometimes collectively referred to as uremic cardiomyopathy, are characterized by LVH, dilated cardiomyopathy, and systolic dysfunction (7- 8). Left ventricular abnormalities are extremely common upon initiation of dialysis, with LVH evident in 40%, LV dilatation in 28%, and systolic dysfunction in 16% of all patients starting dialysis (7).

In spite of extensive knowledge about cardiac structure and function in patients with ESRD, the relationship between mild reductions in renal function and cardiac morphology and function has been less clear. In contrast to the established evidence that ventricular enlargement after MI has been associated with adverse CV outcomes (17- 20), our data suggest that renal impairment was associated with smaller ventricular volumes, while having a significant impact on mortality and HF. Moreover, we observed that lower eGFR was also associated with larger LA volume (Figure 1B) and with the presence of MR in a higher proportion of patients. However, the transmitral Doppler indexes of LV diastolic function did not differ in magnitude among the categories of eGFR. The LA volume is an indirect measure of ventricular pressure and compliance during ventricular diastole (21). The LA volume has been shown to be related to the severity of diastolic dysfunction and to be an important risk factor for development of HF in patients with LV diastolic dysfunction (22- 24).

In our study, declining renal function was also strongly associated with a higher proportion of patients presenting with LVH (60% for eGFR <45 vs. 19% for eGFR ≥75) and with a higher proportion of patients presenting with a history of hypertension (75% for eGFR <45 vs. 48.9% for eGFR ≥75) (Table 1). Left ventricular hypertrophy is likely to develop in the earliest phases of all forms of renal disease with a trend toward higher LV mass with declining eGFR (25). Although the pathogenesis of LVH in patients with renal failure is less clear, hypertension is believed to be a leading cause of LVH in ESRD, and anemia, along with volume overload, is thought to play a potential contributory role (26). Increased myocardial mass leads to abnormal passive elastic properties of the myocardium and myocardial stiffness, reducing chamber compliance and elevating diastolic pressure, which accounts for diastolic dysfunction (27- 28). Diastolic filling abnormalities are common in hypertension, and diastolic dysfunction is believed to be an important pathophysiologic intermediate between hypertension and HF development. (29). An increase in intercardiomyocyte fibrosis and myocardial collagen content (30- 31), coupled with the alterations in calcium homeostasis interfering with myocardial relaxation (32), is also thought to contribute to LV diastolic dysfunction in patients with renal dysfunction.

The increased incidence of CV deaths or hospital stays for HF in patients with poor renal function could be partially explained by an increase in LV mass. Our observation of similar baseline EF and infarct segment length among different categories of renal function suggests that the higher risk of mortality and hospital stay for HF in patients with poor renal function cannot be attributed to consequences of larger infarcts. Greater baseline LA volumes, LV mass, and a higher LV mass/LV EDV ratio with smaller ventricular volumes in patients with lower eGFR in this study suggests that diastolic dysfunction—most likely preceding the MI—could possibly explain the higher risk of CV mortality and HF development in patients with declining renal function.

Several limitations of this analysis should be noted. First, we used the MDRD equation for the estimation of GFR. This is an indirect creatinine-based assessment of renal function, and although considered reliable in estimating GFR, it has limitations, because serum creatinine is influenced by non-renal factors. The MDRD equation might be less accurate in populations without chronic kidney disease, such as young patients with type 1 diabetes without microalbuminuria and in potential kidney donors (33). Second, echocardiograms were obtained, on average, 5 days after MI and as many as 10 days after MI; thus, our baseline echocardiogram might have occurred after substantial recovery of function or after significant ventricular enlargement, both of which can occur during this time period (34). Third, although transmitral Doppler indexes in our study did not differ in magnitude among the categories of eGFR, these are poor measures of diastolic function that are highly sensitive to preload and can change unpredictably as diastolic dysfunction progresses (21,35). More robust noninvasive measures of diastolic function, such as tissue Doppler imaging, were unavailable in the majority of centers participating in the VALIANT study at the time the study began. Finally, our results are predominantly applicable to the high-risk cohort of the VALIANT study. We cannot comment on the potential influence of renal impairment as assessed by eGFR in patients with normal LV function after MI, and neither of these findings can be extrapolated to patients with severe renal dysfunction, because patients with a baseline serum creatinine of >2.5 mg/dl (221 μmol/l) and worse degrees of renal function were not included in the VALIANT study.

In summary, the results of this study suggest that in a high-risk cohort of individuals with LV dysfunction after MI, patients with renal impairment are similar to those without renal impairment with respect to global left and right ventricular systolic function and infarct segment length. Yet those with impaired renal function have smaller ventricular volumes, increased LV mass and mass/volume ratio, and larger LA volumes. Thus, reduced systolic function alone cannot account for increased CV mortality and hospital stay for HF in post-MI patients with impaired renal function. Increased LA volume and LV mass in patients with impaired renal function suggests that diastolic dysfunction might be an important mediator of increased risk of HF development and CV mortality in patients with renal impairment after MI.

Online Cardiosource Slide Set