Stroke and Renal Dysfunction

Abstract

Cerebrovascular disease may represent an important clinical presentation of atherosclerosis in chronic kidney disease (CKD), and atherosclerosis is frequently encountered in CKD. In fact, kidney disease is now considered a risk factor for development of cardiovascular disease. Although guidelines for primary prevention of stroke have been recently published, CKD is hardly mentioned. Based on a series of available studies, we analyzed the relationship between reduced renal function, end-stage renal disease (ESRD), and stroke. Reduced renal function and risk of stroke appear to be related to the highest risk of patients on dialysis treatment. Primary and secondary prevention of stroke should be encouraged in participants with renal dysfunction.

Introduction

Stroke is the third cause of death and the first cause of severe disability in elderly patients in the Western world, ¹ representing a major medical and social problems. ^1,2 In particular, older stroke patients show an increased mortality and more disabling stroke for a similar spectrum of stroke severity and pathology. ³ The presumed stroke etiology should be evaluated as cardioembolic, small vessel, large vessel, or other/undetermined according to the modified Trial of Org 1072 Acute Stroke Treatment (TOAST) criteria. ⁴ In participants having a stroke, the risk of death has been associated with age, ⁵ altered levels of consciousness, ⁶ severity of neurological deficits, ⁷ congestive heart failure, ⁸ atrial fibrillation, ⁹ and diabetes mellitus. ¹⁰ Many biochemical markers, such as C-reactive protein, glucose levels at the time of admission, erythrocyte sedimentation rate, leukocyte count, fibrinogen, uric acid, and low triiodothyronine level have been related to mortality in participants with stroke. ^11–13

The recently published guidelines for primary prevention of stroke provide an overview of the evidence on established and emerging risk factors for stroke. ¹⁴ However, although reduced renal function is now considered a well-established risk factor for cardiovascular disease even in people with and without manifest vascular disease, ^15,16 the American Heart Association (AHA)/American Stroke Association guidelines quoted renal disease only when mentioning the blood pressure (BP) goal to reach in patients with diabetes or kidney disease (<130/80 mm Hg).

Reduced renal function may reflect both the duration and severity of different cardiovascular risk factors such as hypertension, diabetes mellitus, and dyslipidemia, and it is often linked with the development of other less established vascular risk factors such as anemia, oxidative stress, electrolyte imbalance, and hyperhomocysteinemia. ¹⁷ Diagnosis of chronic kidney disease (CKD) is higher among patients with cardiovascular disease than in the general population, and CKD is an independent risk factor for cardiovascular morbidity and mortality. ¹⁸ Early detection of CKD could stimulate its treatment geared toward reducing the deterioration of renal function and preventing the risk of cardiovascular complications. ¹⁹ The incidence of stroke begins to double every decade after the age of 55. ²⁰ Cerebrovascular disease has been recognized as a frequent complication in CKD, even if this relationship needs to be defined more precisely. Therefore, based on a series of recent studies, we reviewed the current knowledge relating stroke to renal function.

Modifiable Risk Factors for Stroke in Renal Dysfunction

Risk factors for stroke could be classified as (a) modifiable, such as hypertension, smoking, diabetes, cardiac disease, and alcohol and (b) nonmodifiable, including age, male, gender, family history of stroke, and nonwhite race. In particular, hypertension plays a key role in the development of stroke events. ^14,21 Dyslipidemia is a well-known risk factor for cardiovascular disease in the general population. In CKD, although the atherosclerotic burden is high, the pathophysiology of dyslipidemia is complex, since renal patients should exhibit lower high-density lipoprotein cholesterol (HDLc) levels, higher triglycerides, and normal or even low low-density lipoprotein cholesterol (LDLc). ²² Nevertheless, also in renal patients, dyslipidemia represents a modifiable risk factor, and statins appear to be an effective treatment. ²² A growing body of evidence has recently shown that venous thromboembolism and artery vessel diseases represent 2 faces of the same coin, sharing common risk factors. ^23,24 Thus, it is not surprising that thromboembolism, atherosclerosis, inflammation, and cardioembolism are the main causes of ischemic stroke in patients with CKD. ²⁵ Hypertension, diabetes, alcohol habits, retinal artery sclerosis, and age were found as independent risk factors for subcortical brain infarction, and incidence of stroke was much more higher in participants with subcortical silent brain infarction compared with those without previous silent stroke. ²⁶ Accelerated atherosclerosis could be the major risk factor for stroke in patients with end-stage renal disease (ESRD). Moreover, ESRD is associated with hypertension, bleeding diathesis, and anticoagulation therapy with heparin during hemodialysis, all factors potentially capable of increasing the risk of hemorrhagic stroke. Again, well-established stroke risk factors such as diabetes, hypertension, cardiac disease, and prior stroke are more frequent in the uremic than in the general population. ^27,28 Thus, patients with CKD are at risk of both thrombotic and bleeding events at the same time. ²⁹ On one hand, in these patients, abnormalities in coagulation and platelet function could act as thrombotic factors in untreated participants and could increase the risk of bleeding in participants on antiplatelet or anticoagulant therapy. ^30,31 On the other hand, the prevalence of atrial fibrillation, a well-known risk factor for development of fatal and nonfatal cerebrovascular events, has been estimated to range ^32–34 between 3% and 23.5%.

Taken together, these data strengthen the concept that CKD represents a particularly vulnerable condition of high cardiovascular risk.

Urinary Abnormalities and Stroke

Studies on autopsy data of patients with stroke found that renal artery stenosis is a common finding. ³⁵ The atherosclerotic process could damage both brain and kidney in a similar way since they have similar vascular structure, with low-resistance exposure of the small vessels to high pulsatile flow and pressure. ³⁶ Again, the most common cardiovascular risk factors, such as hypertension and diabetes, can lead to similar microvascular damage to both the brain and the kidney; so that reduced glomerular filtration rate (GFR) and asymptomatic or symptomatic brain infarcts and white matter lesions are the main clinical manifestations of a systemic disease. ^37–39 Viehman et al, aiming to evaluate the usefulness of urinalysis in discriminating cardioembolic from noncardioembolic stroke, investigated 341 patients with ischemic stroke and observed that urine white blood cells >14.5/µL and urine red blood cells >41.7/µL were predictor of cardioembolic stroke. ⁴⁰ As for urinalysis, however, proteinuria remains the cornerstone. In 1999, Beamer et al reported that microalbuminuria was a common finding in patients with cerebrovascular disease, being associated with increased risk of stroke, although they only investigated 186 participants and predictors of microalbuminuria were diabetes, serum albumin levels, age, and ischemic heart disease. ⁴¹ A few years later, the relationship between proteinuria and stroke was confirmed in cohorts of patients from United Kingdom (microalbuminuria was associated with 50% increased risk of stroke) ⁴² and Japan. ⁴³ A recent meta-analysis, conducted on 46 638 participants who had 11 479 stroke events, was aimed to estimate the risk of stroke associated with microalbuminuria and macroalbuminuria. Incident stroke risk was greater for macroalbuminuria than microalbuminuria and increased proportionally with rising urinary albumin excretion. ⁴⁴ One study has also showed that in patients admitted with ischemic stroke, proteinuria was associated with disability at discharge and with decreasing odds of being discharged home directly from the hospital. ⁴⁵

On one hand, detection of microalbuminuria has to be taken into account since it is strongly and independently associated with incident stroke risk. On the other hand, however, impact of microalbuminuria correction on the incidence of stroke is unknown, and future studies are required in order to explore whether microalbuminuria can be considered a modifiable risk factor. ⁴⁶

Chronic Kidney Disease and Stroke

Since late 1990s, mild renal dysfunction was associated with increased risk of cerebrovascular events among patients with cardiovascular disease, ⁴⁷ and it has also been observed that moderate renal dysfunction was associated with a 37% increased risk of dementia among 3349 participants in the cardiovascular Health Cognition Study. ⁴⁸ Serum creatinine predicted mortality during 18 months of follow-up in elderly stroke survivors, ⁴⁹ and reduced renal function on admission was related to mortality in patients with acute stroke. ⁵⁰ However, a retrospective study investigating risk factors for short-term mortality in 469 older participants with acute ischemic stroke found that mortality was associated with the presence of altered levels of consciousness, congestive heart failure, and age. Serum creatinine was not different in survivors and deceased (1.11 ± 0.65 vs 1.17 ± 0.42 mg/dL). ⁵¹ In a population of 1977 individuals living in the rural Japanese community of Ohasama, Nakayama et al calculated a relative hazard for first symptomatic stroke of 1.9 for creatinine clearance of 40 to 70 mL/min and 3.1 for creatinine clearance <40 mL/min. In contrast, macroalbuminuria was not related to first symptomatic stroke. ⁵² Koren-Morag et al evaluated the incidence of ischemic stroke or transient ischemic attack (TIA) in 6685 patients with chronic coronary heart disease, followed-up for 4.8 to 8.1 years. Patients with CKD exhibited 1.54-fold hazard ratios of incident ischemic stroke or TIA ⁵³ , and CKD has been also related to small-vessel cerebrovascular disease. ³⁹

Given the possible biases of the estimation of kidney function based on serum creatinine evaluation (especially in older participants), GFR estimation represents a more reliable indicator of renal dysfunction. In fact, also the European Society of Hypertension and European Society of Cardiology Task Force on the management of arterial hypertension recommends the use of equations for GFR estimation or the measurement of creatinine clearance as routine laboratory tests. ⁵⁴ In the last couple of years, a growing body of evidence has been collected both under 2 points of view: (a) cardio- and cerebrovascular events in participants with low renal function, (b) renal function in participants with stroke, and (c) the possible impact on clinical outcome.

Cardio- and Cerebrovascular Events in Participants With Low Renal Function

Dalrymple et al compared the rates of ESRD, cardiovascular, and noncardiovascular death in 1268 participants aged 75 ± 6 with GFR <60 mL/min per 1.73 m ² , followed-up for 9.7 years. They concluded that older adults with CKD are 13-fold more likely to die from any cause and are 6-fold more likely to die from cardiovascular disease than progress to ESRD. ⁵⁵ Di Angelantonio et al evaluated the relationship between CKD and risk of major cardiovascular events in 16958 participants aged 33 to 81, followed up for 24 years. Chronic kidney disease was diagnosed in 7% of cases and their mean GFR was 58.7 ± 14.8 mL/min per 1.73 m ² . The results showed that in people without manifest vascular disease, even the earliest stages of CKD were associated with excess risk of subsequent coronary heart disease, although the hazard ratios for death due to ischemic, hemorrhagic, and unclassified stroke were not significant. It has to be observed, ⁵⁶ however, that only 1016 patients (6%) had GFR <60 mL/min per 1.73 m ² and only 12 (0.07%) <30 mL/min per 1.73 m ² . Recently, our group evaluated a cohort of about 200 consecutive patients who presented to the Emergency Medical Service for symptoms not related to cerebrovascular function and renal function, who were assessed for risk of first stroke according to the AHA guidelines. ⁵⁷ We found a mean 10-year stroke risk of 26.5% (±19.8%), with an evident and significant relationship with the stage of CKD. ⁵⁷

Renal Function in Participants With Stroke

In a cohort of 414 participants with stroke, out of 5624 participants aged 55 or older (National Health and Nutrition Examination Survey [NHANES] 1999-2004), stroke survivors were older and more likely to have hypertension, diabetes mellitus, elevated glycohemoglobin and homocysteine levels, and urinary-albumin-to-creatinine ratio as well as lower GFR, hematocrit, and cholesterol concentration compared with participants without stroke. Glomerular filtration rate <60 mL/min per 1.73 m² and urinary-albumin-to-creatinine ratio >30 mg per 1 g of creatinine were independently associated with stroke. ⁵⁸ Tsagalis et al investigated a consecutive series of 1193 first-ever stroke patients followed up for 10 years. Participants were classified into 3 groups: GFR >60, between 60 and 30, and <30 mL/min per 1.73 m², with exclusion of patients with acute kidney injury. Only 30 patients had GFR <30 mL/min per 1.73 m²: they were older, had higher prevalence of hypertension and atrial fibrillation histories, and had most severe stroke compared with the other groups. Participants with GFR <60 mL/min per 1.73 m² had higher mortality rate, and renal function, age, history of TIAs, coronary artery disease, atrial fibrillation, and glucose on admission were independent predictors of the overall composite cardiovascular events during the follow-up. The probability of having a composite cardiovascular event over the follow-up was 45.2, 67.2, and 77.6 in participants with GFR >60, between 60 and 30, and <30 mL/min per 1.73 m², respectively. ⁵⁹ Mostofsky et al evaluated 1175 consecutive patients hospitalized for acute ischemic stroke and examined the shape and magnitude of the relationship between renal function and mortality. During the follow-up (median 40.3 months), 43% of participants were reported to be dead. Among patients with acute ischemic stroke, a reduced or highly elevated GFR at hospital admission was associated with a higher mortality rate compared with patients with moderate levels of GFR (between 75 and 125 mL/min per 1.73 m ² ). The results suggested a U-shaped relationship between GFR and mortality. The authors explained the elevated risk of death in stroke patient with GFR >125 mL/min per 1.73 m ² by hypothesizing a false elevation of GFR due to low muscle mass and concluded that the severity of impaired kidney function in patients hospitalized with acute ischemic stroke is associated with increased all-cause mortality independent of age, sex, race, and major comorbidities. ⁶⁰ Lee et al performed a meta-analysis (21 studies, 284 672 patients with 7863 strokes) to evaluate the relationship between reduced GFR and incident stroke. Incident stroke was higher in participants with reduced GFR, with a relative risk of 1.43 (95% confidence interval [CI] 1.31-1.57) and 1.77 (95% CI 1.32-2.38) in patients with GFR <60 and <40 mL/min per 1.73 m ² , respectively. ⁶¹

As for acute stroke and thrombolysis, Agrawal et al performed a retrospective observational study on 20 patients with CKD who received intravenous tissue plasminogen activator (tPA) for acute stroke. Reduced renal function was not associated with asymptomatic and symptomatic intracranial hemorrhage, poor functional status, or in-hospital death outcomes, suggesting that the use of thrombolytics in acute stroke is appropriate in patients with renal dysfunction. ⁶²

Low Renal Function and Impact on Clinical Outcome

Patients admitted to acute stroke unit with both renal dysfunction and heart failure had highest risk of mortality and onset of new cardiovascular events, with respective hazard ratios of 3.42 and 3.59. ⁶³ In 425 known stroke patients of age 55 or more with baseline history of stroke, although the initial stages of CKD were predictor of future mortality, worsening renal function was related to death, with the highest hazard ratio of death (5.93) for participants ⁶⁴ in stages 4 to 5. Again, people with known stroke and low GFR had higher hazard ratio of dying and the risk increased throughout the different stages of CKD, being higher ⁶⁴ in stages 4 and 5.

On one hand, renal dysfunction could represent a marker of the burden of vascular atherosclerotic damage and could identify participants with very high global cardiovascular risk. Preventive therapeutic strategies should be addressed to this group of participants, and the use of antihypertensive drugs and statins should be encouraged. On the other hand, in patients with acute cerebrovascular event, renal function should be carefully investigated (with GFR formulae and not limiting to serum creatinine levels) and taken into consideration due to the relationship with the risk of mortality. Finally, although the paucity of available data, it seems that patients with stroke and reduced renal function could be taken into consideration for thrombolytic option.

End-Stage Renal Disease and Stroke

Cerebrovascular disease is highly prevalent in hemodialysis patients; the prevalence is much higher than in the general population. A retrospective study evaluating 1064 ESRD patients found that cerebrovascular events developed in 9.2% of cases. The incidences of cerebral hemorrhage and infarction were 8.7 and 3.7 per 100 patient-year, respectively. ⁶⁵ Moreover, the frequency of silent stroke is 5 times higher in hemodialysis patients than in the general population. ⁶⁶ In particular, the incidence of hemorrhagic stroke is high in patients with CKD and dialysis patients, ^65,67,68 and it seems to be particularly relevant in the Japanese hemodialysis populations. ^65,68–70 Seliger et al performed a population cohort study comparing rates of hospitalized ischemic and hemorrhagic stroke among incident dialysis patients recorded in the US Renal Data System database and the National Hospital Discharge Survey. ⁷¹ Unadjusted stroke hospitalization rate for dialysis patients was 479.4 per 10 000 person-year, markedly higher than that of the general population (29 per 10 000 person-year). The age-standardized risk of stroke hospitalization was calculated to be 6.1 for Caucasian and 4.4 for African American among males with ESRD. The relative risk among females with ESRD was 9.7 for Caucasian and 6.2 for African American when compared with the general population, indicating that the risk was higher in Caucasian than African American, and in women. The excess of stroke risk was reported for both ischemic and hemorrhagic stroke, being, however, higher for ischemic stroke. ⁷¹ Again, the same authors tried to identify, with a multivariate model of analysis, patient characteristics associated with the risk of hospitalized or fatal stroke in dialysis patients in a cohort of 8920 dialysis patients (77% of whom with complete data available). ⁷² The incidence of stroke was 33/1000 person-years, confirming a very high rate of stroke in ESRD. Markers of malnutrition were strong predictors of incident stroke, as was mean BP (11% increased risk associated with an increase of 10 mm Hg). Incidence of hemorrhagic stroke was 4.6/1000 person-years with higher mean BP associated with higher risk. Incidence of ischemic stroke was 25.2/1000 person-years. ⁷² Sozio et al evaluated the incident dialysis patients in a prospective national cohort study (CHOICE Study; 81 clinics, 1041 patients enrolled 1995-1998, followed-up until December 2004). ⁷³ In all, 165 participants (15.9%) had incident cerebrovascular event after dialysis therapy initiation, and they were more likely to be of older age, male sex, have lower diastolic BP, history of diabetes mellitus, prior cerebrovascular disease, or prior peripheral arterial disease at baseline. Seventy-six percent of the total event were ischemic; 35% of all events were fatal (28% of ischemic and 90% of hemorrhagic strokes). Overall, 56% of the patients who had a stroke were able to be discharged to home or acute rehabilitation, with the best recovery seen in those with small-vessel occlusion (78% discharge rate), and the worst outcome in those with hemorrhagic stroke (10% discharged to home or acute rehabilitation).

Length of time on dialysis treatment could account for the difference in the different studies. In fact, it is likely that ischemic stroke is more frequent in the early months of dialysis treatment, while long-term dialysis treatment could be related to acquired risk factors for hemorrhagic stroke. In the study of Sozio et al, ⁷³ in fact, the incidence rate for cerebrovascular events was greatest in the first 2 years after dialysis therapy initiation. On one hand, the reduction in BP values due to ultrafiltration could diminish cerebral perfusion, leading to ischemic stroke. On the other hand, bleeding diathesis and routine heparin administration during hemodialysis could increase the risk of hemorrhagic stroke.

Antithrombotic and/or anticoagulant therapy might play a role, also considering the concomitant presence of atrial fibrillation. The international Dialysis Outcomes and Practice Patterns Study (DOPPS) showed that the use of aspirin in dialysis patients was associated with a reduced risk of stroke, but the strength of this association was modest. ⁷⁴ Chou et al conducted a retrospective longitudinal study evaluating 219 dialysis patients with atrial fibrillation who had been followed up for 36.9 ± 21.9 months. Fatal and nonfatal cerebrovascular events developed in 22.4% of participants with permanent atrial fibrillation, 39.5% of those with paroxysmal atrial fibrillation, and 32.6% of those in whom paroxysmal atrial fibrillation transformed to permanent atrial fibrillation. ⁷⁵ In contrast, Wieshlzer et al calculated that the annual risk of stroke in hemodialysis patients was 1%; however, atrial fibrillation was not a risk factor for stroke. ⁷⁶

Stroke is common in dialysis patients, with a predilection for hemorrhagic events. It seems that length of time on dialysis treatment could play a role in the different stroke subtype etiology, with ischemic events more frequent in the early months of dialysis, and hemorrhagic strokes more likely to occur in the course of long-term dialysis treatment, may be also due to the development of acquired hemorrhagic risk factors.

Conclusions

Taken together, these data suggest that patients with CKD are predisposed to stroke, especially as the GFR decreases. Thus, low renal function deserves to be considered in the list of stroke risk factors. Although we cannot state that CKD is a direct cause of stroke, there are no doubts that a strong association between these 2 conditions exists. Brain and kidney appear to be closely linked, just from the earlier stages of CKD, as shown by the association between microalbuminuria and stroke. Although we do not know, at least now, whether correction of proteinuria may have an impact on stroke incidence, detection of microalbuminuria should be recommended since it is strongly and independently associated with incident stroke risk. Renal dysfunction can be probably considered a marker of the burden of vascular atherosclerotic damage, linked with very high global cardiovascular risk, and GFR should be carefully estimated since it could be an useful tool to detect early stages of CKD in patients with or without overt coronary heart disease, who might benefit from lipid-lowering therapy. ²² Preventive therapeutic strategies should be addressed to patients with low renal function, especially based on the use of antihypertensive medications, antiplatelet agents, and anticoagulants. ⁷⁷ Statins could also represent an effective treatment. Many guidelines around the world recommend statin prescription in patients with CKD. According to the 2009 Canadian Cardiovascular Society/Canadian guidelines, ⁷⁸ in high-risk individuals, hypolipidemic treatment should be started immediately, concomitant with health behavior interventions with respect to appropriate diet, physical activity, weight management, and smoke cessation. The primary target of therapy is to achieve a 50% reduction in LDLc from baseline values (class I, level A). The majority of patients will be able to achieve target LDLc levels on statin monotherapy. However, a significant minority of patients may require combination therapy with an agent that inhibits cholesterol absorption (ezetimibe), which can also decrease LDLc by an additional 20% for ezetimibe. A recent randomized placebo-controlled trial study aimed to assess the efficacy (and safety) of the combination simvastatin plus ezetimibe in a large cohort of patients with CKD (4650 treated vs 4320 untreated patients) showed that this combination treatment safely reduced the incidence of major atherosclerotic events (−17%) in a wide range of patients with advanced CKD. ⁷⁹ Finally, particular attention should to be addressed to dialysis patients, remembering that duration of dialysis treatment could play a role in the different stroke subtype etiology, for example, ischemic or hemorrhagic.