Cardio–renal organ damage and cardiovascular outcomes in hypertension

Introduction Hypertensive organ damage is currently regarded as an intermediate stage in the continuum of cardiovascular disease and as a strong determinant of total cardiovascular risk; reversal of organ damage represents a useful intermediate end-point for assessing the efficacy of antihypertensive treatment. A growing body of evidence underlines the relevance of organ damage as a determinant of cardiovascular morbidity and mortality; indeed, quantitative markers of organ damage [i.e. increased left ventricular mass, carotid intima–media thickening, reduced glomerular filtration rate (GFR) and urinary albumin excretion] have been associated with an increased incidence of cardiovascular events independently of traditional risk factors. In this issue of the Journal, Tsioufis et al.[1] report the findings of a prospective study aimed at comparing the prognostic value of left ventricular hypertrophy (LVH) and chronic kidney disease (CKD) in predicting coronary artery disease (CAD), stroke, all-cause mortality and the composite end-point in a cohort of 1652 uncomplicated essential hypertensive patients attending an outpatient hypertension clinic in Greece. The relationship of major traditional risk factors, including age, sex, current smoking, diabetes mellitus and dyslipidaemia, with incident cardiovascular events and mortality is also addressed. The authors show that echocardiographic LVH, defined as LV mass indexed to body surface area equal to or exceeding 125 g/m2 in men and 110 g/m2 in women, according to 2007 European Society of Hypertension/European Society of Cardiology (ESH/ESC) guidelines [2] and CKD diagnosed on the basis of estimated GFR ranging from 15 to 59 ml/min/1.73 m2 are both significant predictors of the composite end-point; moreover, LVH but not CKD is an independent predictor for stroke. Despite some methodological limitations that deserve a critical analysis, this paper offers a new piece of information on the key role of cardiac and renal damage either alone or combined in stratifying cardiovascular risk in the hypertensive setting. Before addressing this issue in detail, some more general considerations on the available evidence in this area might be useful. Left ventricular hypertrophy and cardiovascular prognosis LVH is a marker of cardiac damage which reflects a long-term exposure to pressure overload as well as to multiple cardiovascular risk factors and may identify individuals more exposed to the adverse effects of high blood pressure (BP). Several studies have underlined the relationship between LVH at baseline examination, either assessed by electrocardiography (EKG) or by the more accurate echocardiographic method, and morbidity/mortality in population-based samples and in hypertensive cohorts. Despite the limited diagnostic accuracy of EKG criteria, a robust body of evidence has been accumulated in the past decades about their capability to predict adverse cardiovascular outcomes. In a recent study including 7495 adult Americans of different ethnicity (24.4% African–Americans, 51.7% whites and 23.9% Latinos) examined in the National Health and Nutrition Examination Survey (NHANES) III study, a new EKG indicator of LVH, the Novacode estimate of LV mass index, has been reported to be significantly associated with 10-year cardiovascular mortality in the three racial groups [3]. This relation remained significant after adjusting for age, systolic BP, smoking, cholesterol and diabetes, with a hazard ratio significantly greater for African–Americans [2.3, 95% confidence interval (CI), 1.55–3.42] than for whites (1.32, 95% CI 1.14–1.76) and Latinos (2.11, 95% CI 1.35–3.30). LV mass, calculated according to validated formulae based on echocardiographic measurements of LV thickness and diameter, has been shown to be a predictor of nonfatal cardiac and cerebrovascular events as well as of cardiovascular and total mortality in a variety of settings; this was quite independent from diagnostic thresholds, type of indexation to body size and sex-specific criteria adopted for LVH definition [4]. Notably, echocardiographic studies reporting a relationship between LVH at baseline and subsequent cardiovascular events as well as a favourable effect of LVH regression are all based, with the exception of one [5], on calculated LV mass. The risk of cardiovascular morbid events, cardiovascular death and all-cause mortality is 2–8-fold higher in hypertensive individuals with LV mass above normal limits compared with those with normal LV mass [6]. In addition to excellent accuracy in identifying LVH, quantitative echocardiography provides a variety of information on LV structure and geometry, systolic and diastolic function as well as coexistent cardiac alterations. Prospective studies indicate a stepwise increase in cardiovascular morbidity and mortality with LV mass increments, from a low event rate in individuals with normal LV mass and geometry, to intermediate rates in those with eccentric LVH, and to the highest level of risk in patients with concentric LVH. The gradient in cardiovascular risk in patients with different LVH geometric patterns is related to several factors, including cardiac mass itself, cardiac function and systemic haemodynamic profiles. An updated study [7] including a large population of normotensive and hypertensive participants (n = 35 602) with normal ejection fraction (≥50%) identified an abnormal LV geometry in 46% of patients, LV concentric remodelling representing 35% (n = 12 362) and LVH 11% (n = 3958) of cases. A strong relation was found between abnormal LV geometry and mortality; in particular, patients with LV concentric remodelling and LVH exhibited a consistently higher relative risk for all-cause mortality compared with participants with normal LV geometry (relative risk 1.99, 95% CI 1.88–2.18, P < 0.001 and 2.13, 95% CI 1.89–2.40, P < 0.0001, respectively). Moreover, regression from concentric remodelling to normal LV geometry was associated with an improved survival compared with progression to LVH. Data from a population-based study conducted on a cohort of 1772 African–Americans free from CAD showed that the relationship between LV patterns and all-cause mortality was markedly attenuated after adjustment for baseline differences in cardiovascular risk factors and LV mass in women, but not in men, in whom the association between concentric LVH and mortality risk remained significant [8]. The clinical implication of reversal of concentric LVH has been recently documented by Muiesan et al.[9] who showed that changes from concentric to normal (or eccentric) geometry have a favourable prognostic value. Overall, both EKG and echocardiographic studies [10,11] performed in population-based samples, on high-risk individuals and on hypertensive cohorts, have clearly documented that LVH regression significantly decreases cardiovascular event risk, whereas its persistence or progression implies a worse prognosis, independently of BP control and treatment modality. In the Losartan Intervention for Endpoint Reduction in Hypertension (LIFE) echocardiographic substudy [12], the only prospective randomized trial investigating the prognostic value of LV mass modification, a 25 g/m2 decrease in LV mass index (i.e. one standard deviation) was associated with a 22% reduction in the composite end-point of cardiovascular morbidity and mortality during 4.6 years of follow-up. Chronic kidney disease: epidemiology and cardiovascular prognosis Accumulating evidence shows that CKD is a clinical condition strongly associated with subclinical cardiovascular alterations and is a potent predictor of cardiovascular events. Indeed, manifestations of CKD such as renal insufficiency and microalbuminuria or macroalbuminuria are associated with a higher prevalence of cardiovascular disorder and predict adverse cardiovascular outcomes in different clinical settings [13]. An accurate estimation of kidney function is fundamental for CKD detection and management. Renal insufficiency is preferentially classified according to the estimated GFR calculated by the abbreviated Modification of Diet in Renal Disease (MDRD) formula that requires age, sex, race and serum creatinine. Estimated GFR values below 60 ml/min/1.73 m2 indicate stage 3 CKD, whereas values below 30 and 15 ml/min/1.73 m2 define stage 4 and 5 CKD, respectively. The Cockcroft–Gault formula estimates creatinine clearance based on age, sex, body weight and serum creatinine values; this formula is valid for GFR greater than 60 ml/min, as it tends to overestimate creatinine clearance in CKD stage 3–5 as well as in overweight and obese individuals. CKD is an increasingly emerging phenotype; its prevalence, however, remains underestimated in current practice due to the limited implementation of the disease criteria as provided by guidelines [2,14]. In a survey of a large representative sample of US adults, performed between 1999 and 2000, the prevalence of moderate-to-severe kidney dysfunction defined as estimated GFR of 15–59 ml/min/1.73 m2 was 4.4%, whereas the prevalence of mildly deranged kidney function was 36.3% (estimated GFR ranging from 60 to 89 ml/min/1.73 m2) corresponding to stage 2 CKD, according to the Kidney Disease Outcome Quality Initiative Clinical Practice guidelines [15]. In a study performed in eastern Ontario [16], renal function data from 17 independent outpatient laboratories were retrospectively analysed to determine the proportion of adults with at least one serum creatinine determination during a 1-year period and the prevalence rates of low GFR according to different algorithms. Of the 1 090 000 residents, 32% had at least one serum creatinine determination during the study year. Sixteen percent of the population (49% of those in whom creatinine determination was performed) had at least one GFR less than 80 ml/min/1.73 m2 and 5% (16%) had at least one GFR less than 60 ml/min/1.73 m2. The prevalence of renal dysfunction in patients with long-standing hypertension is significantly greater as compared with that found in community-based studies, varying with the setting; for instance, an estimated creatinine clearance less than 60 ml/min was found in a consistent fraction (14%) of a cohort of 787 essential hypertensive patients attending a hospital clinic [17]. Studies on selected populations, such as middle-aged men, individuals at high risk for cardiovascular disease, patients with recent stroke or myocardial infarction and predialysis patients have consistently shown an association between CKD (i.e. reduced renal function) and cardiovascular events. Most population-based studies, with only few exceptions, have confirmed such a detrimental association. In the Cardiovascular Health study [18], serum creatinine levels greater than 133 μmol/l were associated with a 73% increase in all-cause mortality risk. In a population-based cohort of 631 participants followed up for 10.2 years, renal function was inversely related to all-cause and cardiovascular mortality, regardless of whether renal function was estimated by serum creatinine, the Cockcroft–Gault formula or Levey's prediction equation [19]. A decrease in GFR of 5 ml/min/1.73 m2 was associated with a 26% increased risk of cardiovascular death. A large body of evidence is also available about the prognostic significance of renal dysfunction in hypertension. In the Hypertension Optimal Treatment (HOT) study, a trial including 18 790 hypertensive patients, aimed at comparing the incidence of cardiovascular events according to three diastolic BP targets (≤90 mmHg, ≤85 mmHg, ≤80 mmHg), renal insufficiency at baseline, as indicated by a serum creatinine concentration more than 1.5 mg/dl or an estimated creatinine clearance less than 60 ml/min, was accompanied by a significant elevation of cardiovascular risk [20]. In particular, patients with an estimated creatinine clearance less than 60 ml/min showed significantly higher adjusted risk ratios for cardiovascular and all-cause mortality compared with those with values more than 60 ml/min (1.80, CI 1.33–2.44, P < 0.001; 1.65, CI 1.34–2.03, P < 0.001), respectively. Results from the Syst-Eur trial, a randomized, placebo-controlled, double-blind intervention study on 4695 elderly patients with isolated systolic hypertension [21], indicated that serum creatinine and proteinuria are independent indicators of total mortality and all cardiovascular end-points. Furthermore, analysis of data according to GFR rather than serum creatinine revealed that GFR was a more powerful predictor for all end-points. For instance, for decline in GFR of 20 ml/min, cardiovascular mortality rose by 60% and the risk for any cardiovascular complications by 48%. The association of renal function, as assessed by Cockcroft–Gault-estimated creatinine clearance, and ischaemic heart disease mortality has been investigated in a large, multiethnic cohort of 9929 treated hypertensive patients [22]. After adjusting for major risk factors, a progressive decrease in GFR was associated with an increased risk of ischaemic heart disease mortality: hazard ratios ranged from a 41% increased risk when estimated GFR was 60–74.9 ml/min/1.73 m2 to nearly a three-fold increased risk for estimated GFR less than 45 ml/min/1.73 m2. Ruilope et al.[23] have recently compared estimated creatinine clearance [Cockcroft-Gault (GC) formula] with estimated GFR (MDRD formula) as predictors of cardiovascular outcome in 15 245 high-risk hypertensive participants in the Valsartan Antihypertensive Long-Term Use Evaluation (VALUE) trial. Estimated creatinine clearance less than 60/ml/min was significantly predictive of all-cause death (hazard ratio = 1.22, 95% CI 1.076–1.390, P < 0.0021), whereas estimated GFR was predictive of all outcomes except stroke: primary cardiac end-point 1.497 (1.332–1.682, P < 0.0001), myocardial infarction, 1.501 (1.254–1.796, P < 0.0001), congestive heart failure, 1.699 (1.435–2.013, P < 0.0001) and all-cause death, 1.231 (1.098–1.380, P < 0.0004). These results suggest that, at least in the hypertensive setting, GFR calculated by the MDRD formula is a more sensitive tool than estimated creatinine clearance in predicting cardiovascular outcomes. Microalbuminuria is a marker of impaired renal function and, more generally, of endothelial dysfunction; it is an established risk factor for cardiovascular morbidity and mortality and for end-stage renal disease in individuals with a high cardiovascular risk profile (i.e. diabetes mellitus). Increasing evidence, however, derived from prospective observational trials indicates that urinary albumin excretion well below current microalbuminuria thresholds is also associated with an increased risk of cardiovascular disease and all-cause mortality [24]. Even in healthy normotensive, nondiabetic individuals such an association has been documented. Microalbuminuria is present in about 4–8% of apparently healthy individuals without any significant cardiovascular risk factor. Several studies suggest that prevalence of microalbuminuria is greatest in populations with both hypertension and diabetes. In the International Survey Evaluating Microalbuminuria Routinely by Cardiologists in Patients with Hypertension (i-SEARCH) study [25], an international survey, which enrolled 21 050 hypertensive patients from 26 countries, the prevalence of microalbuminuria, as assessed by one-time dipstick, was 58.4%, with considerable variation across countries (from 53 to 71%). Independent correlates of microalbuminuria were male sex, metabolic syndrome, diabetes mellitus, congestive heart failure and CAD. Rates of microalbuminuria observed in i-SEARCH exceed the prevalence rates previously reported in the hypertensive setting (ranging from 4 to 46%) as well as in a primary healthcare large-scale study performed in Germany (19%) [26]. New insights and limitations of the paper by Tsioufis et al. The report by Tsioufis et al. [1] confirms and expands previous evidence on the prognostic relevance of hypertensive organ damage by showing that echocardiographic LVH and reduced GFR are both significant predictors of all-cause death and cardiovascular morbidity. As prespecified primary and secondary outcomes were compared by stratifying individuals into four groups according to the presence of LVH and CKD [LVH(−)–CKD(−), LVH(+)–CKD(−), LVH(−)–CKD(+), LVH(+)–CKD(+)], the study provides a novel piece of information on the separate as well as on the combined value of subclinical cardiac and renal damage for predicting cardiovascular events. The group with both cardiac and renal damage at baseline examination, as compared with that without any organ damage (reference group), exhibited a 2.5-fold greater incidence of CAD (P = 0.003), 4.0-fold increased risk for stroke (P = 0.002) and 3.2-fold higher risk for the composite end-point (P < 0.001). Interestingly, no differences in all studied end-points were found between group LVH(+)–CKD(−) and group LVH(−)–CKD(+). In a multivariate Cox regression analysis, after inclusion in the model of a number of covariates such as age, sex, smoking, diabetes and dyslipidaemia, age more than 65 years, male sex, LVH and CKD turned out to be independent predictors of the composite end-point, whereas only age and LVH were independently associated with incident stroke. These findings deserve to be briefly commented on. First, data from the present study indicate that combined cardiac and renal involvement defines a very high-risk phenotype affected by worse prognosis than any single damage. Due to the importance of multiple organ damage in determining cardiovascular prognosis in the hypertensive population, this condition should be carefully sought in order to optimize therapeutic interventions. In a study [27] carried out in a large cohort of 2213 treated essential hypertensive patients attending an outpatient hospital hypertension clinic, we have recently shown that a clustering of two or three markers of organ damage was more prevalent than a single organ involvement (46.4 vs. 36.4%). The frequency of multiple organ damage in our series was greater than that reported by Tsioufis et al. in the present study, in which the combination of LVH and CKD encompassed less than 10% of the total population. This large difference may be ascribed to the fact that our analysis included additional markers of organ damage such as microalbuminuria and carotid atherosclerosis. Second, LVH but not CKD was found by Tsioufis et al. to be a powerful predictor for subsequent cerebrovascular events, supporting the idea that LVH detection may be useful in identifying individuals at high risk of incident stroke and in planning an aggressive prevention. Indeed, a recent meta-analysis has reported that LVH regression is a strong surrogate marker of decreased risk for cerebrovascular complications in hypertensive patients [28]. An interesting contribution of the present study refers to the findings obtained in supplementary analyses based on alternative definitions of CKD and LVH. The predictive value of CKD defined by the Cockcroft–Gault formula was similar to that derived from the MDRD equation: this result does not confirm the superiority of this latter approach, as shown in the VALUE trial, in the prediction of cardiovascular outcomes in essential hypertension. Notably, when LVH was defined according to LV mass normalized to height2.7, by using the partition values derived from the Lang's intersociety report (i.e. 49 g/m2.7 in men and 45 g/m2.7 in women), its prevalence rose to 60.5% as compared with 35.6% originally detected by the criteria recommended by the 2007 ESH/ESC guidelines [2]. This is because indexation to height2.7 has been shown to provide the highest prevalence of LVH, as it identifies both obesity-related and BP-related LVH [29]. Thus, normalization of LV mass to height2.7 enhanced LVH capacity to predict CAD incidence; in fact, by traditional indexation to body surface, a large proportion of overweight and obese patients (likely to be affected by major metabolic disturbances) with increased LV mass were mistakenly categorized in the group without cardiac damage. Some limitations of this study should also be acknowledged. First, an important criticism is that no albuminuria data are available on this population; this may have add some information on the relationship of this marker with GFR, LVH and cardiovascular outcomes. Second, definition of the CKD phenotype was based on a single determination of serum creatinine for estimation of GFR; this may have misclassified renal status in patients with transient rather than chronic renal dysfunction. Indeed, results from a large Canadian database have shown that low GFR values at a first determination were confirmed in subsequent testing performed within at least 30 days in only 68% of the individuals surveyed [16]. Finally, as recognized by the authors, although the observed relationship between cardio–renal damage and cardiovascular prognosis remained significant after adjustment for traditional risk factors, they did not account for the emerging risk factors, in particular highly sensitive C-reactive protein. Notwithstanding, the data provided by Tsioufis et al.[1] show that parallel cardiac and renal involvement in hypertensive individuals without overt cardiovascular disease is associated with a very high risk of future cardiovascular events and mortality; additional prospective studies, based on a more accurate detection of the of cardio–renal are to the prevalence and the prognostic of this detrimental in different clinical The has no of