Statistics from Altmetric.com
- cardiac arrhythmia
- renal dialysis
- sudden cardiac death
- implantable cardioverter-defibrillator
- atrial arrhythmias
- atrioventricular block
- renal disease
Chronic kidney disease (CKD) is defined as evidence of kidney damage or a glomerular filtration rate (GFR) ≤60 ml/min/1.73 m2 (table 1). The most common causes of CKD are hypertension and diabetes mellitus. The many causes of CKD are associated with different varying prognoses. Patients with adult polycystic kidney disease have a 50% lifetime risk of needing dialysis compared with 25% for type 1 diabetes and <5% for type 2 diabetes. Dialysis is usually considered when GFR falls below 10 ml/min/1.73 m2 but the exact timing will often be dictated by clinical circumstances. This may be refractory oedema, hyperkalaemia and acidosis, uraemia or unacceptable symptoms. Dialysis only partially replaces the excretory function of the kidneys and so the morbidity and mortality associated with CKD are not completely resolved with dialysis. In fact, mortality in the dialysis patient is very high. The life expectancy of a 25-year-old dialysis patient is 12 years, compared with 32 years for an age equivalent transplant recipient and 52 years for a 25-year-old in the general population.1 Even patients with CKD stage 5 will only have a 20–25% chance of surviving long enough to require dialysis. The greatest cause of death in CKD is premature cardiovascular disease. For example, fewer than one in five patients with heart failure will have a normal GFR and 38% of the prevalent dialysis population have coronary artery disease (CAD) (17% previous myocardial infarction, 23% symptoms of angina).2 Both cardiac and renal systems appear to be completely interdependent, further emphasising the concept of the ‘cardiorenal syndrome’. This is highlighted when considering arrhythmias in patients with impaired renal function.
The arrhythmia burden of the patient with CKD is high, with the single greatest contributor to mortality in end stage renal disease (ESRD) being sudden cardiac death (SCD). SCD accounts for 25% of all cause mortality in dialysis patients, at a rate of 10–25% per year depending on comorbid factors.3 Atrial fibrillation may be present in one in three dialysis patients and is associated with an 80% 5 year mortality.4 Furthermore, both ventricular and supraventricular arrhythmias can be caused by haemodialysis. The absence of sinus rhythm in the diabetic patient on haemodialysis has been shown to confer a dreadful prognosis, with an 89% increased risk of death and 75% increased risk of cardiac death or myocardial infarction.5 There are a number of pathophysiological factors associated with CKD that are likely to contribute to this, and also factors associated with CKD that have a direct impact on the management of arrhythmias in this population.
The patient with CKD may progress through a number of key stages of their disease process and management, conferring different physiological challenges that may provoke arrhythmias. The initial phase will be progression of renal disease that ultimately requires renal replacement therapy (RRT). RRT is most likely to take the form of haemodialysis, but may be peritoneal dialysis. Both of these pose many pathophysiological challenges to the patient with some unique factors associated with haemodialysis. Some patients will ultimately have renal transplantation where many of these challenges are removed and some pathophysiological developments may resolve. However, this is associated with different challenges in the form of immunosuppression. Some patients will ultimately return to RRT with dialysis in the event of transplant failure, and hence the cycle begins again.
CAD is likely to be at least in part responsible for arrhythmias in CKD. There are a number of reasons why CKD patients will have an increased risk of CAD (figure 1). The pathological processes that cause CAD are also likely to lead to CKD. For example, chronic hyperglycaemia and insulin resistance in diabetes mellitus can cause endothelial dysfunction and subsequent atheroma formation in all vasculature. It is likely that the proteinuria seen in diabetic nephropathy, a marker of adverse outcome, is an early vascular effect of endothelial dysfunction.6 Patients on dialysis have a two- to fivefold fold increase in coronary artery calcification compared to the general population and this is reflected in the high prevalence of CAD.7 However, the vascular calcification is a diffuse phenomenon in CKD and is associated with hyperphosphataemia. There is a suggestion that phosphate and other unmeasured metabolites induce osteoblast-like activity in vascular smooth muscle. The resultant calcification means that vessels lose compliance and elasticity. Hence, there is an inability to adapt to changes in blood pressure, particularly hypotension. Vascular calcification is associated with adverse cardiovascular outcome and so may contribute to SCD. In the presence of one or both of autonomic dysfunction and rapid fluid shift associated with dialysis, it is clear why vascular calcification may cause circulatory collapse.
Cardiac autonomic neuropathy may also contribute to arrhythmia. Both diabetes and uraemia are potential causes of autonomic neuropathy. Diabetic autonomic neuropathy is in part a vascular phenomenon, but at a microvascular level. Hyperglycaemia causes neuronal vasoconstriction and oxidative stress which results in nerve ischaemia. Diabetic autonomic neuropathy is associated with potential electrocardiographic markers of risk of SCD, such as heart rate variability. However, these alone are not proven to be independent markers of risk.
CKD is also independently associated with endothelial dysfunction. Some pro-inflammatory cytokines and chemotactic factors are renally excreted but are not removed during dialysis. Hence, there is a state of chronic leucocyte activation. ‘Uraemia’ is a marker of a multitude of oxidative and pro-inflammatory metabolites that are not routinely measured but that are likely to contribute to this. Biochemical tests of inflammation such as C reactive protein are associated with adverse cardiovascular outcome in CKD. Chronic inflammatory processes may explain the increased mortality in patients who dialyse via centrally tunnelled venous catheters when compared to arteriovenous fistulae, which carry a much lower risk of infection.
While poor left ventricular function is relatively common in the dialysis population, it is not prevalent enough to completely account for the high rate of SCD. Echocardiography studies in dialysis patients have shown that structural abnormalities are common. Dilated cardiomyopathy is seen in 36% of dialysis patients and left ventricular hypertrophy (LVH) in 74%.8 These are both associated with dialysis induced arrhythmia. The majority of haemodialysis patients will be hypertensive. The combination of this and volume overload promotes LVH. Endomyocardial biopsies in dialysis patients with dilated cardiomyopathy show abnormal remodelling with interstitial fibrosis and myocyte hypertrophy. It is likely that these processes promote arrhythmias, as LVH is associated with a 60% increased risk of SCD in diabetic dialysis patients.5 In addition, other pathological aspects of CKD, such as anaemia, uraemia, and hyperparathyroidism, independently contribute to myocardial fibrosis.
Haemodialysis is associated with dramatic changes in electrolytes with progressive changes between dialysis sessions and rapid changes during dialysis. The potential for arrhythmias associated with electrolyte changes and the chronic, often profound electrolyte abnormalities associated with CKD are well established. Hyperkalaemia (table 2) is the most obvious of these, but hypokalaemia is very common and associated with dietary restrictions, loop diuretics, and the low potassium content of dialysis fluid. The optimum pre-dialysis serum potassium concentration is 4.6–5.3 mmol/l, and using both too high and too low dialysate potassium are associated with intra-dialytic arrhythmias.9 Both hyper- and hypokalaemia may precipitate life threatening arrhythmias.
Hypocalcaemia occurs in most dialysis patients. The inability of the failing kidney to produce 1-hydroxy-vitamin D results in poor dietary absorption of calcium. Arrhythmias rarely occur in the setting of hypocalcaemia alone as parasthesiae and muscular tetany will usually occur at a serum calcium level before cardiac risk is high, indicating the need for intervention.
The prevalence of atrial fibrillation (AF) in pre-dialysis CKD patients is between 9–21%. This increases to between 13–27% in patients on long term haemodialysis. One in three dialysis patients will have experienced a supraventricular arrhythmia.10 This most commonly occurs in the hours immediately after dialysis and is associated with dialysis induced ischaemia and abnormal pre-dialysis potassium. Ten per cent to 15% will suffer AF independently of dialysis sessions. AF in these circumstances is associated with an 80% 5 year mortality.4 However, the arrhythmia itself is not an independent predictor of death, rather an indicator of underlying cardiovascular disease. LVH and dilatation, coronary artery and valvular diseases are associated with AF in haemodialysis patients, and are all more common in these patients than the general population. Atrial flutter has been shown to occur as often in dialysis patients as AF. Hyperparathyroidism and LVH are more common in patients with either AF or flutter when compared to other dialysis patients.11 The impact of atrial flutter on mortality in these patients is not known. Other tachycardias, such as atrioventricular re-entry (AVRT), atrioventricular nodal re-entry (AVNRT), and atrial tachycardias are also less well reported in CKD and are not associated with dialysis.
Antiarrhythmic drug treatment (ADT) is often approached with caution in ESRD because of the fear of adverse drug effects. However, dose adjustments are not often indicated (table 3). For example, there is no evidence that the adverse effects of β-blockade increase significantly with decline in GFR. While this indicates that maintenance doses need not usually be changed for patients with CKD, it is vital to understand that the presence of CKD affects the transport and distribution of a drug as well as its metabolism. Therefore, signs of potential adverse effects of these drugs should be actively sought on a regular basis.
The relative effectiveness of ADT in CKD is not well understood. Management should follow conventional therapy, with any necessary dose adjustments being made. Catheter ablation can be effective in this setting and is an appropriate strategy in difficult to manage patients, but will occasionally be complicated by the vascular access difficulties commonly associated with the haemodialysis population. Another clinical challenge is the use of anticoagulant therapy for AF in dialysis patients. There is an apparent excess of cerebrovascular events in anticoagulated dialysis patients, but warfarin treatment in CKD stage 5 has never been prospectively studied. As with so many drugs used in CKD, the assumption of therapeutic benefit is extrapolated from studies which largely exclude dialysis patients. Its use should therefore be decided on an individual case basis. Figure 2 outlines the key points in the management of supraventricular tachycardia.
Ventricular tachyarrhythmias and SCD
In the general population, the rate of SCD is 1 event per 1000 patient years. In CKD, the risk increases by a hazard ratio of 1.1 for every 10 ml/min decline in eGFR, with an overall event rate of up to 7.8 per 1000 patient years.12 This rate increases 10-fold for patients starting dialysis and increases further with time on dialysis.3
United States Renal Data System (USRDS) data show an increase in SCD of 50% on Mondays and Tuesdays for haemodialysis patients. This reflects the first dialysis activity after a 2 day weekend dialysis break.13 This may reflect repolarisation instability from electrolyte change, poor homeostatic adjustment, sudden fluid removal or an ischaemic effect of dialysis. However, a haemodialysis patient is most likely to die suddenly in the last 12 h of the long weekend break between dialysis sessions, suggesting build up of fluid and electrolytes rather than their removal as the cause.
Intra-dialytic cardiac arrest occurs at a rate of 7 per 100 000 dialysis sessions, with an equal occurrence of brady- and tachyarrhythmias.14 In comparison to peritoneal dialysis patients, haemodialysis patients are more likely to suffer SCD in the first 6 months after starting a dialysis programme. The rates merge for patients established on either modality for 6 months or more.3 This may reflect an early peak of sudden death in patients who are prone to arrhythmic death as a consequence of haemodialysis, or more simply that patients with multiple comorbidities are usually less able to self-care on peritoneal dialysis and therefore undertake haemodialysis.
Medical management of arrhythmias in patients with CKD should focus on risk reduction for SCD and can be divided into drug treatments, patient lifestyle modifications, and changes in dialysis prescription, where relevant. Randomised trials of drug treatments almost universally exclude dialysis patients, and often patients with anything more than CKD stage 2 or 3. As a consequence there is no evidence based guidance on the medical management of arrhythmias in CKD. Minimising the risk of SCD in CKD patients currently focuses on general cardiovascular risk reduction. Tight control of blood pressure, fluid balance, serum phosphate, anaemia, and avoidance of hypo- and hyperkalaemia are all associated with cardiovascular survival benefit.
Cardioselective and non-selective β-blockers appear to be safe with no need for dose reduction. Although there have been no large randomised trials to this effect, β-blockade would seem to be advisable as a first line antihypertensive agent in advanced CKD and dialysis patients because of a definite reduction in cardiovascular mortality and an inferred reduction in the risk of SCD.15 This benefit may be due to the reduction in LVH and slower progression of heart failure associated with β-blocker use in these patients as much as a direct antiarrhythmic effect. The use of β-blockers does not increase the risk of hypotension compared to other antihypertensive agents.
Implantable cardioverter defibrillator therapy
The use of implantable cardioverter defibrillators (ICDs) in dialysis patients is not common practice in spite of their recommendation in international guidelines.16 These recommend the use of ICDs in patients with life threatening ventricular arrhythmias, especially in those awaiting renal transplantation. The guidelines further qualify that patients should be receiving optimal medical treatment and have a reasonable expectation of survival with a good functional status for more than a year. However, the use of ICDs in this population remains controversial, as the upfront cost of ICDs is expensive in a population that already utilises significant health resources. All cause mortality is significantly higher than in the general population and so survival benefit from ICDs may be less. It is perhaps for this reason that the landmark trials of ICD therapy all excluded patients with advanced CKD. There are a number of retrospective observational studies on the use of ICDs in the CKD population but no comprehensive prospective evaluation. The presence of CKD independently predicts poor outcome in patients with ICDs,17 though a clear survival advantage over medical treatment exists, with a 42% risk reduction in patients with ICDs reported in one study. This relative risk reduction is comparable to that seen in the general population treated with ICDs for secondary prevention.18
Given the multifactorial mechanism for SCD in dialysis patients, it is understandable that few dialysis patients who suffer SCD would have fulfilled the criteria for ICD implantation according to current guidelines. Seventeen per cent of SCD occurs in patients with a left ventricular ejection fraction <30% with a 5 year SCD risk of 60%. The lifetime risk of SCD for dialysis patients with normal left ventricular function is 25%.19
While there may be a role for ICDs in primary prevention in the dialysis population, close attention should be focused on modifying risk factors. The pathophysiological factors that contribute to sudden death in ESRD are, therefore, likely to be multiple and heterogeneous compared to post-myocardial infarction patients in whom the substrate for arrhythmias and SCD is myocardial scar. This explains why risk stratification is clinically challenging in this population in spite of the high rate of SCD. It is possible that modification of risk early in CKD may be a potentially more effective approach to primary prevention than device therapy. Figure 3 outlines the key points in the management of ventricular tachycardia.
Bradycardia is common in diabetic patients on dialysis and may be more pronounced with the concomitant use of α- and β-blockers and some calcium channel blockers. Diabetes and uraemia are associated with cardiac autonomic neuropathy. This can be reflected in abnormal electrocardiographic markers such as heart rate variability. This may be a tool to aid risk stratification.
The sudden fluid shifts seen on dialysis often result in hypotension and a secondary tachycardia. However, a minority of patients will suffer bradycardia that is an effect of cardiac under-filling due to hypovolaemia rather than a paradoxical response to blood pressure change. The primary arrhythmia in these patients is still usually sinus tachycardia, but volume depletion eventually leads to a sudden loss of sympathetic tone. Management of these patients should focus on careful assessment of fluid status and a review of medication. Potassium may also be implicated as bradycardia is often seen in severe hyperkalaemia. Bradycardia appears to be a common malignant rhythm in sudden death on dialysis units, though is not widely reported in the literature. Further evaluation of the arrhythmia burden of this population is required. Profound bradycardia as a consequence of a cardio-inhibitory response seen in vasovagal syndromes is no more common in dialysis patients than in healthy controls, suggesting that this is one rhythm abnormality not directly associated with CKD.
The use of pacing devices in dialysis patients is associated with up to a 50% complication rate. These include early infections, sensing/capture abnormalities, and vascular access complications. Seventy per cent of patients with ipsilateral tunnelled dialysis lines and pacing devices will develop stenoses.20 However, there is no apparent associated worsening of survival or indication to withhold pacemaker device therapy in the CKD population.
Arrhythmias in the renal transplant patient
SCD is uncommon in renal transplant recipients, the rate falling somewhere between the general population and CKD patients. Anaemia, LVH, blood pressure, fluid balance, and electrolyte control all improve after transplantation. This patient group is usually on many drugs not encountered in routine clinical practice. The interactions of some of these drugs have the potential to provoke arrhythmias (table 4). The polypharmacy of dialysis and transplant patients is not mutually exclusive. As a transplant fails and a patient crosses back over to dialysis, he or she will need to be treated as both a dialysis and transplant patient and arrhythmic risk managed accordingly. This also highlights the fact that a CKD patient's GFR will fall with time, and so what may have been a safe stable dose of a drug can produce adverse events when there is significant or unexpected loss of renal function. This transition from being a failing transplant patient to re-starting dialysis is a period of high mortality (18 deaths per 100 patient years, compared to 8 deaths per 100 patient years while on the transplant waiting list). The majority of deaths in this period will be of cardiovascular or infective origin.21 A live donor renal transplantation in the present day may have a life expectancy of 15 years or more. Close monitoring of the polypharmacy in renal transplantation is mandatory.
The risk of arrhythmia associated with CKD is very high and prevention through modification of cardiovascular risk factors should be of priority. Supraventricular arrhythmia is associated with a poor prognosis, largely as a reflection of underlying heart disease. Antiarrhythmic therapies are largely safe in these patients provided that there is diligent monitoring for potential adverse effects.
The risk of ventricular arrhythmias and SCD appears to be very significant, though poorly studied. There is a significant survival benefit from ICD therapy and international guidelines advocate their use in dialysis patients. Therefore, ICDs should be considered where appropriate. Device therapy may yet prove to be of some benefit to other CKD patients, but universal primary prevention with β-blockade may be more viable and effective. It is apparent that the high incidence of arrhythmia and SCD is the end point of a unique, chronic, and catastrophic cardiovascular burden that CKD carries, and highlights the need for cohesive, joint cardiorenal management of these patients.
Arrhythmias in chronic kidney disease: key points
β-blockers are advised, but sotalol should be avoided.
Drugs which are not affected by renal metabolism may still have an altered distribution or binding in chronic kidney disease (CKD).
All drug treatment must be closely monitored and possible interactions sought thoroughly.
Haemodialysis induced arrhythmias alone do not necessarily confer a poor prognosis.
Sudden cardiac death (SCD) is the most common cause of death in end stage renal disease, but there is little evidence informing its management.
Current guidelines do not exclude implantable cardioverter defibrillator use in CKD patients.
CKD and post-myocardial infarction patients are not comparable in the nature of their SCD risk.
You can get CPD/CME credits for Education in Heart
Education in Heart articles are accredited by both the UK Royal College of Physicians (London) and the European Board for Accreditation in Cardiology—you need to answer the accompanying multiple choice questions (MCQs). To access the questions, click on BMJ Learning: Take this module on BMJ Learning from the content box at the top right and bottom left of the online article. For more information please go to: http://heart.bmj.com/misc/education.dtl
RCP credits: Log your activity in your CPD diary online (http://www.rcplondon.ac.uk/members/CPDdiary/index.asp)—pass mark is 80%.
EBAC credits: Print out and retain the BMJ Learning certificate once you have completed the MCQs—pass mark is 60%. EBAC/ EACCME Credits can now be converted to AMA PRA Category 1 CME Credits and are recognised by all National Accreditation Authorities in Europe (http://www.ebac-cme.org/newsite/?hit=men02).
Please note: The MCQs are hosted on BMJ Learning—the best available learning website for medical professionals from the BMJ Group. If prompted, subscribers must sign into Heart with their journal's username and password. All users must also complete a one-time registration on BMJ Learning and subsequently log in (with a BMJ Learning username and password) on every visit.
Extensive reporting from the US database of patients with renal disease.
Study of >4000 patients with ESRD examining the demographic factors associated with CAD in this specific population. The study identified conventional risk factors for CAD and additional factors associated with uraemia such as low serum albumin.
Study examining routine ECGs in patients with diabetes on dialysis focused on establishing indicators of risk. This study identified a dramatic increase in risk in patients not in sinus rhythm.
Review article of the pathophysiological effects of diabetes and its impact on endothelial function and associated cardiovascular disease.
Review article outlining the value and importance of vascular stiffness in determination of cardiovascular risk in patients with renal disease.
Comprehensive review of the incidence and common clinical challenges of atrial fibrillation in the CKD population. There is a detailed evaluation of the stroke risk and the value of anticoagulants in the context of ESRD.
Study evaluating the association of GFR and risk of SCD in patients with moderate renal disease and known CAD.
Evaluation of the records of 400 witnessed cardiac arrests on dialysis, examining common clinical characteristics of the patients and their outcomes.
Comprehensive guideline document based upon extensive evaluation of the medical literature. Specific section highlighting the management of ventricular arrhythmias in the patient with ESRD.
Competing interests In compliance with EBAC/EACCME guidelines, all authors participating in Education in Heart have disclosed potential conflicts of interest that might cause a bias in the article. The authors have no competing interests.
Provenance and peer review Not commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.