15b Cardiovascular Problems in Chronic Kidney Disease

Until recently, impaired kidney function was an orphan in the cardiological assessment of patients with heart disease, while in patients with chronic kidney disease (CKD) cardiological assessment was not considered a high priority. However, impressive evidence has now been provided that cardiovascular problems critically determine life expectancy of patients with CKD.

Cardiological examination of patients with advanced CKD should be obligatory. Importantly, invasive diagnostic workup and revascularization procedures should not be withheld in CKD patients presenting with myocardial ischaemia.

Conversely, in cardiac patients appropriate assessment of renal function employing modern methodology (estimated glomerular filtration rate, albuminuria) is obligatory as well, since kidney function critically determines systemic neurohumoral activation.

In the management of renal patients the most important aspect is appropriate control of blood pressure, but cardiovascular prevention in these patients comprises the full spectrum of beta blockade, blockade of the renin–angiotensin–aldosterone system, aspirin, and statins.

It has been recognized only recently that minor CKD is a powerful cardiovascular risk factor [1]. The recent Consensus Group therefore stated that renal evaluation should be part and parcel of the evaluation of patients with cardiac problems [2].

The assessment of renal function is based primarily on two parameters, the glomerular filtration rate (GFR) on the one hand and urinary excretion of albumin (or in more advanced stages of CKD unselective proteinuria) on the other hand.

For the clinical assessment of GFR, one has to be aware of the fact that in early stages of CKD a major reduction of GFR may still be compatible with serum creatinine concentrations within the normal range, because the serum creatinine concentration depends apart from the GFR also on non-renal factors, particularly on muscle mass. This frequently causes underestimation of the severity of CKD in the elderly and cachectic patient. This dilemma has led to efforts to provide more accurate estimates of GFR (eGFR) by standardizing the measurement of serum creatinine (Cleveland clinic protocol) and by using an algorithm correcting for age, gender, and ethnicity:

[3].

For eGFR values >60mL/min/1.73m², the estimate is imprecise and the laboratory should state only that the value is >60mL/min/1.73m². It is known, however, from studies in large cohorts that already minor reductions of eGFR in the range of values above 60mL/min increase the cardiovascular risk [4]. One of the reasons for the insensitivity of GFR to detect incipient kidney damage is the fact that the remaining nephrons respond to nephron loss with compensatory hyperfiltration, thus masking the extent of renal damage by initially still maintaining the whole kidney GFR within the normal range. Importantly, in the normal and near-normal range of GFR cystatin C is more precise than creatinine based eGFR. The measurement, however, is costly and currently not (yet) used for routine measurements.

Currently the severity of chronic kidney disease is graded from CKD1–CKD5 ( Fig. 15b.1).

Apart from GFR, urinary excretion of albumin is a powerful independent predictor of cardiovascular risk: the risk increases progressively with rising urinary albumin concentrations. For historic reasons, one still distinguishes between normoalbuminuria and microalbuminuria (definition of microalbuminuria: excretion of 30–300mg albumin/day), but the cardiovascular risk steadily increases

even in the range of normoalbuminuria. The most convenient and adequately sensitive procedure is to measure albumin in the morning urine without correction for urinary creatinine concentration.

It is important to note that both GFR and albuminuria independently contribute to the cardiovascular risk [5]. In the absence of albuminuria, the cardiovascular risk of reduced renal function is markedly less than if a combination of low GFR plus albuminuria is present ( Fig. 15b.2).

The impact of impaired renal function on the cardiovascular risk is not only relevant in patients with primary kidney disease, but is also pronounced in patients with primary cardiac disease, particularly in patients with the acute coronary syndrome (ACS). Fig. 15b.3 shows that in patients with ACS the entire spectrum of complications is progressively more frequent the more advanced the CKD [6] is. CKD is also a powerful predictor of risk in patients with congestive heart failure.

A most important part of the assessment in renal malfunction is the evaluation of blood pressure. In renal patients this apparently simple problem becomes problematic. Table 15b.1 makes this point: office blood pressure is notoriously unreliable. Many studies showed that self-measured blood pressure is a much more reliable parameter. Renal patients, particularly diabetic patients with nephropathy, are characterized in early stages of CKD by nocturnal non-dipping which makes ambulatory blood pressure measurement a very valuable tool.

Vascular stiffening is a main consequence of impaired renal function. This changes vascular impedance and contributes to large blood pressure amplitudes. In renal patients, central blood pressure (in the aorta) is usually considerably higher than brachial blood pressure, thus complicating the

issue of target blood pressures on treatment. Because coronary perfusion occurs only during diastole, it appears wise not to lower diastolic blood pressure to values <70mmHg [7], at least in patients with known coronary heart disease. To halt progression of CKD, current guidelines recommend a target blood pressure of 130/80mmHg (or lower values if proteinuria exceeds 1g/day) in patients with diabetic and non-diabetic CKD [8]. Renin–angiotensin system (RAS) blockade further intensifies reduction of proteinuria independent of blood pressure [8]. Since reducing albuminuria has a significant impact on cardiovascular events [9], one should consider intensifying RAS blockade to reduce proteinuria even to values <1g/day, if necessary by using doses of RAS blockers beyond those licensed for blood pressure-lowering.

It has been recognized that in patients with primary kidney disease the risk to ultimately require haemodialysis is much lower than to die mainly from cardiovascular causes—by a factor of 20 (CKD2) in early, and a factor of three in late stages of CKD (CKD4) [10]. This clearly highlights the importance of diagnosing renal dysfunction in early stages of CKD in order to obtain a complete assessment of the cardiovascular risk in patients with primary kidney disease. Due to the systemic impact of kidney failure on, amongst others, neurohumoral activation, detection of CKD is especially relevant in patients with primary heart disease.

The mechanisms by which renal impairment affects cardiac function are not entirely clear, but the most important aspects for the management of patients are the early appearance of:

The prospective 4D study [13] in haemodialysed type 2 diabetic patients showed that cardiac arrest (26% of deaths), heart failure (6%), and death from other cardiac causes (3%) are more frequent causes than myocardial infarction (9%), although myocardial infarction is much more frequent than in the background population. This finding has been confirmed by the US Renal Data System Registry in non-diabetic patients as well.

Left ventricular systolic dysfunction with a low ejection fraction at echocardiography is well known as a strong predictor of cardiovascular survival [14]. The importance of diastolic malfunction, however, a typical finding early on in kidney failure, has been vastly underestimated in the past.

Accelerated large vessel disease with increased arterial stiffening and calcification leads to an increased afterload to the heart inducing left ventricular hypertrophy, which in turn is greatly amplified independent of blood pressure, since kidney malfunction affects the heart by (currently poorly characterized) factors other than volume and pressure. The resulting uremic cardiomyopathy is mainly characterized by cardiomyocyte drop-out with interstitial fibrosis (demonstrable with magnetic resonance imaging, MRI), by right and left ventricular hypertrophy, and by microvessel disease with wall thickening of post-coronary arteries and capillary deficit leading to a marked decrease of ischaemia tolerance. In addition, our experimental studies documented diminished insulin-dependent glucose uptake. Importantly, the pro-arrhythmic mechanisms triggered in this molecular concert of hypertrophy encompass changes on various levels of the organ ranging from autonomic failure down to dysfunctional subcellular Ca²⁺ handling, all these components paving the way for triggered activity and re-entry mechanisms as the basis of ventricular tachycardias.

The loss of the ‘Windkessel’ effect causes an increase of wave propagation velocity imposing an increased pulsatile pressure and flow to the peripheral vasculatures, which again induces small vessel damage and end organ dysfunction. Furthermore the resulting greater blood pressure amplitude with low diastolic pressures jeopardizes coronary perfusion.

To assess large vessel disease, an important contributor to the cardiovascular risk, it is useful to have pelvic and abdominal X-rays to assess calcification of iliac arteries and the aorta (both of which are powerful predictors of cardiovascular risk).

Cardiological workup of the renal patient initially includes electrocardiogram and echocardiographic evaluation to detect inappropriate left ventricular hypertrophy and pericardial effusion. Because in end-stage renal disease metabolic myopathy often precludes a meaningful evaluation by treadmill testing, pharmacologic stress testing has become the gold standard in non-invasive detection of cardiac ischaemia. MRI technology has recently allowed the differentiation of two major types of uremic cardiomyopathy, one with a more subendocardial late enhancement possibly reflecting silent ischaemia, and another one with a diffuse pattern of late enhancement characteristically found in patients with more pronounced left ventricular hypertrophy. Moreover, MRI is a powerful tool to detect cardiac amyloidosis. Unfortunately gadolinium application is critical in end-stage renal disease due to its toxicity (see Acute reactions and nephrogenic skin fibrosis after gadolinium, p.514).

The coronary angiogram represents the gold standard for the diagnosis of coronary artery disease (CAD) allowing direct treatment. It is still underused in renal patients because of concern about radiocontrast nephropathy, but should definitely not be omitted in patients with an ACS because cardiac mortality significantly outweighs the potential further reduction of GFR. Observational data (USRDS) show that in dialysed patients percutaneous intervention (PCI) with stenting results in superior short-term outcome, while bypass surgery, using internal mammary grafts, provides superior 1-year survival [15, 16]. In the future, hybrid approaches employing a combination of minimally invasive direct coronary artery bypass grafting (MIDCAB) and PCI might prove beneficial especially in elderly multimorbid renal patients.

Acute renal failure (or according to recent nomenclature: acute kidney injury (AKI)) after radiocontrast administration is a complication in CKD patients, particularly diabetic CKD patients.

Patients with pre-existing CKD account for 30–35% of patients with acute renal failure in general and in such patients dialysis independence at day 90 is less than in patients without CKD. The long-term outcome is also worse in patients with acute renal failure who had pre-existing CKD: 28.2% after 3 years vs. 7.6% of patients without pre-existing CKD. These long-term consequences add to the problem of increased hospitalization and cost due to radiocontrast nephropathy. Several studies show a significantly increased mortality with odds ratios up to 5–10, while an increased rate of delayed death was also shown after radiocontrast-induced acute renal failure during percutaneous coronary intervention [17].

Recent studies show that measurement of indicators of renal damage (kidney injury molecule-1 (KIM-1), neutrophil gelatinase-associated lipocalin (NGAL) and others) [18] allow the recognition of high renal risk within hours of exposure to an intervention, e.g. cardiac surgery; they promise to become clinically useful tools. Whether they also predict radiocontrast nephropathy is unknown.

The first step to prevent contrast nephropathy is to identify patients at higher risk: the elderly and the patients with diabetes, with elevated serum creatinine, and with hypertension. Drugs which should best be avoided in patients scheduled for radiocontrast investigation are non-steroidal anti-inflammatory agents (both COX-1 and COX-2), aminoglycosides, cyclosporine, tacrolimus, amphothericin, etc. There is some recent information that intensive blockade of the RAS may aggravate the risk of acute renal failure.

A long list of interventions has been proposed for the prevention of radiocontrast nephropathy. The only intervention for which uncontroversial evidence is available is the administration of saline, half normal saline [19] or better, normal saline [20]. There have been proposals to administer N-acetyl cysteine [21] or to use sodium bicarbonate [22] instead of saline, but the efficacy of these interventions has not been consistently confirmed. Prophylactic haemodialysis is also not useful.

A bone of contention is the type of radiocontrast agent administered: it has been reported that the iso-osmolar, dimeric, non-ionic radiocontrast agent iodixanol is superior to the low osmolar, monomeric, non-ionic agent iohexol [23], but again this has not been confirmed by subsequent studies. The roles of radiocontrast ionicity, osmolality, and viscosity in the genesis of radiocontrast nephropathy remain currently unresolved. Therefore, the best advice remains to hydrate the patient with normal saline and to administer the lowest possible dose of radiocontrast agents in high risk patients.

MRI has become a powerful diagnostic tool in cardiology as gadolinium-containing contrast agents (Gd-contrast) allow the sensitive detection of myocardial scarring in various specific cardiac diseases such as ischaemic heart disease, myocarditis, cardiomyopathies, amyloidosis etc. Although Gd-contrast has initially been embraced as a non-toxic contrast material in renal patients, both acute and chronic toxicity have been observed especially in end-stage renal patients on haemodialysis and chronic ambulatory peritoneal dialysis.

Nephrogenic systemic fibrosis (NFS) is a rare but potentially fatal condition first described as a scleromyxoedema-like fibrosing syndrome in association with renal insufficiency [24]. It is initially characterized by red and painful plaques that coincide with oedema. Subsequently thickening, induration, and hardening of the skin in the distal extremities and the trunk occurs, while the face is usually spared. Notably, other organs including the lungs, liver, muscles, and the heart may also be involved causing considerable morbidity and mortality. The underlying mechanism is thought to be transmetallation whereby in exchange for endogenous metals free gadolinium is released from the chelate with subsequent binding to tissue. Circulating cells are then recruited causing fibrosis by cytokine production and T-cell activation. Since no effective therapeutic intervention is available, it is of paramount importance to identify the patients at highest risk for NFS. As the majority of patients with NFS were preterminal or on renal replacement therapy (GFR <30mL/min/1.73m²), it is expert opinion that Gd-contrast should only be administered to patients with CKD1–CKD3. Moreover, in observational studies inflammatory state, metabolic acidosis, high calcium and phosphate levels as well as high erythropoietin (EPO) dosages were aggravating factors. Importantly, the vast majority of NFS cases appeared in patients who received gadodiamide (Omniscan®), a contrast agent with a linear and uncharged molecular configuration. It is thus wise to not only to minimize the dose of contrast agent, but also to use Gd-contrast with cyclic and charged configuration posing the least risk of transmetallation.

Recently in dialysis patients a gadolinium-exposure induced systemic inflammatory response was observed in 13 out of 136 patients receiving Gd-DTPA (Magnevist®) [25]. A peracute septicaemia-like clinical picture evolved with fever, malaise, hypotension, vomiting, and dyspnoea. While steroids did not improve symptoms, significant improvement was seen within the first 5 hours of dialysis. Interestingly, C-reactive protein levels remained markedly elevated up to 14 days. Lymphopenia was seen in all patients, PMN remained normal, and none of these patients developed nephrogenic systemic fibrosis.

A major dilemma is the fact that in the past renal patients were deliberately excluded from major intervention studies. As a result, current therapeutic recommendations are mainly based on observational data or based on post hoc analyses in patients with early stages of CKD who had been included in large cardiological intervention studies. Controlled prospective information is available only for few interventions ( Table 15b.2).

The most important component of treatment is blood pressure control ( target values listed in ‘Chronic kidney disease as a cardiovascular risk factor’, p.511). There is a caveat, however, in patients on haemodialysis. They often develop hypotensive episodes during fluid removal by ultrafiltration. Not only high, but even more potently low blood pressure predicts death in this high-risk population, particularly in the elderly with high comorbidity and low diastolic blood pressure values. The advice is to gradually lower blood pressure close to or within the normal range, but not to tolerate hypotensive episodes.

Statins, if anything, were equally effective or even more effective, in patients with reduced kidney function who had been included in intervention trials. Statins are safe and no excessive frequency of rhabdomylosis is seen in CKD patients. One interventional study, the 4D study in haemodialysed type 2 diabetics, however, failed to document a significant effect on the primary composite endpoint. But adjudicated coronary death was lowered to the same extent as in studies on non-renal patients. Therefore it is widespread opinion that all renal patients should be on statins. Most fibrates are metabolized by the kidney. Because of the risk of rhabdomyolysis they require monitoring and dose adjustment: they are therefore little used.

Observational studies showed that in dialysed patients with baseline haemoglobin (Hb) <10g/dL cardiac function was improved and left ventricular hypertrophy partially reversed when Hb was raised by the administration of EPO. Recent controlled studies showed no significant benefit when Hb concentrations were raised by EPO treatment above the recommended target of 11–12g/dL [26]. There is recent concern about adverse effects when Hb is increased further. Therefore Hb values >13g/dL should be avoided.

Even when GFR is not yet decreased, sympathetic activity is increased in patients with primary kidney disease and hypertension as documented by microneurography. In addition, breakdown of circulating catecholamines is reduced in advanced CKD because catecholamine breakdown by the amino-oxidase renalase from the kidney is diminished. Excessive sympathetic activation provides a good a priori rationale for the use of beta-blockers. The modern beta-blockers (carvedilol, nevibolol) with less metabolic and renal circulatory side effects are advisable [27]. The only prospective evidence of cardiovascular benefit available is one single study [28] in haemodialysed patients with cardiomyopathy and reduced EF. It documented significant reduction of total mortality and cardiovascular death. Nevertheless, not to the least because of the high frequency of sudden death, many experts believe that administration of beta-blockers is indicated unless there are specific contraindications.

Inappropriate activation of the RAS is a hallmark of CKD. The PEACE study [29] documented significantly reduced all-cause mortality in patients with stable coronary heart disease and eGFR <60mL/min/1.73m² who had received 4mg trandolapril, but not in patients with eGFR >60mL/min/1.73m², suggesting increased RAS activity or increased responsiveness to RAS blockade in CKD. Intervention studies in renal patients, usually to prevent progression of CKD, are underpowered to show a significant benefit on CV endpoints and this is true also for one intervention study [30] in haemodialysed patients.

In CKD patients RAS blockade by ACE inhibitors or ARB is also indicated because of the well-documented effect of RAS blockade to reduce progression of CKD. Their administration is safe, specifically also with respect to hyperkalaemia [31].

In dialysis patients, maintenance of residual renal function is associated with better survival. Both ACE inhibitors and ARB prolong the persistence of residual renal function. Therefore their administration is rational even in the absence of controlled evidence.

The safety of additional aldosterone blockade for patients with impaired renal function is currently not well documented and hyperkalaemia is a major concern.

Recently, a large number of observational studies in patients with coronary heart disease and in CKD patients (pre-dialysis and on dialysis) suggest improved survival and less cardiovascular events with administration of the precursor 25(OH)vitamin D₃ and of active vitamin D compounds which obviously have actions beyond bones and mineral metabolism. This area is currently in flux and large studies are ongoing. It is wise to measure 25(OH)D concentrations in CKD patients (measurement of 1,25(OH)₂D₃ is not necessary except in unusual hypercalcaemic cases). Cholecalciferol should be administered if the 25(OH)D level is <30ng/mL. Administration of active vitamin D should currently be based on the guidelines is restricted to lowering elevated PTH, but cardiovascular protection by active vitamin D may become a topic in the future.

Even in non-renal patients, serum-phosphate predicts cardiovascular events [11]. The impact of phosphataemia on survival and coronary heart disease has been grossly underestimated in the past. According to current guidelines in CKD patients, serum-phosphate should be kept within the normal range by restriction of dietary phosphate intake and, if necessary oral phosphate binders (for details see guidelines to be published soon). In dialysed patients serum-phosphate should be <6mg/dL.

Sodium retention and hypervolaemia are common in renal patients. Dietary salt restriction (recommendation: 6g/day) and use of diuretics is therefore an integral part of patient management. The plausibility of this recommendation is heightened by experimental and observational evidence that salt—apart from increasing blood pressure—causes also blood pressure-independent CV target organ damage.

In dialysed patients, the issue of salt restriction is controversial.One important predictor of death is loss of residual renal function, i.e. diuresis. Diuretics increase the amount of urine excreted, but do not extend the duration of residual diuresis. In contrast, RAS blockade maintains residual diuresis for longer periods of time.

With respect to volume control, continuous ambulatory peritoneal dialysis (CAPD) has the advantage of slow low-intensity fluid removal, thus avoiding the large volume swings seen with haemodialysis. An alternative currently under investigation is daily dialysis.

Observational studies showed a strong correlation between serum-homocysteine concentrations and survival on dialysis. One interventional study to lower homocysteine by folate showed no benefit.

These two factors are impressively strong predictors of mortality and adverse cardiovascular events, both in CKD and haemodialysis patients. Unfortunately procedures with proven benefit are currently not available.

The use of diuretics for the treatment of kidney and heart failure is an essential component of the therapeutic regimen both in CKD patients and in patients with heart failure. Diuretics are beneficial by potentiating the pressure lowering effect of inhibitors of the RAS and relieving symptomatic episodes of decompensated heart failure. On the other hand, however, diuretics trigger counterbalancing antinatriuretic mechanisms, e.g. ACE II, aldosterone, decreased systemic blood pressure. Such systemic neurohumoral counterregulation has adverse effects on heart and kidneys. Therefore it is expert opinion that the lowest effective dose should be administered. Even relatively low doses of diuretics allow to uptitrate blood pressure lowering medication, e.g. inhibitors of the RAS and beta-blockers—medications with proven benefit in both heart and kidney failure.

In combination with RAS inhibitors low dose thiazide diuretics augment the reduction of albuminuria—a powerful cardiovascular and renal risk factor. RAS inhibitors diminish the thiazide-induced reduction in GFR. As a result, the filtered sodium load is increased, permitting more effective natriuresis.

In patients with a GFR less than approximately 30mL/min, thiazides cause only a minor increase of natriuresis. At this level of GFR they should be exchanged for, or combined with, loop diuretics which are effective even in advanced kidney failure. The combination makes sense because frequently resistance to loop diuretic monotherapy develops because of a compensatory increase of distal tubular Na⁺ reabsorption; this can be overcome by the addition of a thiazide. It is important that in proteinuric patients the efficacy of diuretics is decreased because in the tubulus lumen diuretics are up to 90% protein bound while only the concentration of the free diuretic inhibits tubular Na⁺ reabsorption. In this situation higher doses of diuretics are required. It is important to pay attention to the half-life of the diuretic; e.g. furosemide is fully effective only with three times daily administration.

Na⁺ reabsorption is also increased by aldosterone. The plasma aldosterone concentration usually decreases initially after administration of ACE-inhibitors and ACE receptor blockers, but in the long term a secondary increase in plasma aldosterone frequently occurs (‘escape’).

In proteinuric CKD this is usually associated with an increase in proteinuria. In this situation spironolactone is helpful, particularly in patients with low K⁺ serum levels. Unfortunately at low GFR the risk of hyperkalaemia is significantly increased.

A frequent cause of insufficient efficacy of diuretic treatment is an excessive dietary sodium intake or administration of Na⁺ by infusions. In such patients even though an adequate diuresis may be achieved the sodium balance may not become negative.

In oedematous patients with advanced heart failure, delayed intestinal absorption of diuretics may occur as a result of mucosal oedema in the gut so that less diuretic is delivered to the tubule accounting for an inadequate natriuretic response. If in such cases the intravenous application of diuretics fails to establish diuresis, more invasive measures e.g. ultrafiltration, haemodialysis, inotropic support, or left assist devices may be needed depending on the prevailing pathophysiology of either renal or cardiac failure ( Table 15b.3).

Particularly in elderly patients, mostly women, thiazides and particularly the combination of thiazides with loop diuretics (‘sequential nephron blockade’) may cause hypovolaemic hyponatraemia. This is usually associated with an excess of antidiuretic hormone (ADH). Discontinuation of thiazide diuretics may be sufficient. Hypovolaemic hyponatraemia predicts poor outcome particularly in end-stage heart failure. The treatment is largely empirical: fluid (water) restriction (<1.5L) and discontinuation of diuretics. Small studies found an increased efficacy from concomitant hypertonic saline infusions. Since the primary problem is water retention from ADH excess, this procedure is not recommended: it may contribute to further volume expansion if diuresis is not adequate [32]. The water retention underlying hyponatraemia in heart failure is the result of unopposed baroreceptor activity causing inappropriate non-osmotic release of arginine vasopressin leading to water retention. The administration of vasopressin V₂ receptor blockers represents a novel therapeutic option in hypovolaemic hyponatraemia, but long-term benefit on survival has not been demonstrated.

Loop diuretics, and to a lesser degree, thiazides increase the renal production of prostaglandins. They act as vasodilators and regulate renal blood flow. Non-steroidal anti-inflammatory drugs impair prostaglandin synthesis and thus cause renal ischaemia by unopposed actions of angiotensin II. They also reduce natriuresis. Therefore it is risky to administer them in patients with CKD and advanced cardiac disease.

For full references and multimedia materials please visit the online version of the book (http://esctextbook.oxfordonline.com).