This section addresses a series of important topics which impact the outcomes of people
with CKD. Specifically CVD, interpretation of tests for CVD, infections, vaccinations
and hospitalizations, medications, and patient safety are addressed. This section
is designed to help the practitioner appreciate the specific nuances related to testing
and drug administration in people with CKD and to alert all clinicians to the fact
that many common tests may not be as useful in people with CKD to diagnose or evaluate
therapies. It was beyond the scope of the guideline and evidence review to thoroughly
evaluate all tests with respect to sensitivity and specificity in different groups
of people with CKD but future research may consider this as a useful direction.
4.1 CKD and CVD
Population-based studies have demonstrated an increased risk of death and cardiovascular
mortality as GFR falls below 60 ml/min/1.73 m2 or when albumin is detected on urinalysis.
This is not explained by an increase in traditional risk factors. There are CKD-specific
risk factors associated with more advanced CKD which drive the high rates of mortality
and morbidity even at young ages. People with CKD are more likely to experience a
cardiovascular event than to progress to ESRD, have a worse prognosis with higher
mortality after acute myocardial infarction (MI), and have a higher risk of recurrent
MI, heart failure and sudden cardiac death. Management of modifiable cardiovascular
risk factors, such as improved BP and diabetes control, also reduces CKD progression.
4.1.1: We recommend that all people with CKD be considered at increased risk for cardiovascular
disease. (1A)
RATIONALE
This statement is worded in this way to reflect the strong and independent associations
between GFR and albuminuria categories and risk of CVD in people with CKD and applies
to both adult and pediatric populations.
Evidence Base
Large cohort studies have demonstrated the strong and independent associations between
CVD (acute coronary syndrome [ACS], stroke, heart failure and sudden cardiac death)
and CKD by category of eGFR, after adjusting for known CVD risk factors, history of
CVD events, and proteinuria. In those with an eGFR of 45-59 ml/min/1.73 m2, risk is
increased by 43% and in those with eGFR below 15 ml/min/1.73 m2, risk is increased
by 343%.58 Although people with GFR category G5 (GFR<15 ml/min/1.73 m2) are at the
highest risk of a CVD event, there will be more events in people with GFR categories
G3a-G3b (GFR 30–59 ml/min/1.73 m2) because of the much higher prevalence at these
categories.
420
These events occur at a younger age in people with CKD suggesting that CKD promotes
CVD at an accelerated rate.
421
The prognosis after an acute event is related to level of GFR with a significant rise
in mortality when eGFR falls below 45 ml/min/1.73 m2.
422, 423, 424
Albuminuria is associated with duration and severity of hypertension; an adverse lipid
profile with higher levels of total cholesterol, triglycerides and lipoprotein(a)
and low HDL-C levels;
425
and abnormalities of coagulation. The presence of higher levels of proteinuria increases
the risk of mortality and MI independently of level of eGFR.
426
Many studies have demonstrated low levels of urinary albumin to be associated with
the increased risk of CVD in people with diabetes independent of renal function; however
population studies of non-diabetic individuals have confirmed that even small amounts
of albuminuria are associated with increased CVD risk. In the Third Copenhagen study,
in people with microalbuminuria, risk of coronary heart disease was increased independently
of age, sex, renal function, diabetes, hypertension, and plasma lipids.
427
The Chronic Kidney Disease Prognosis Consortium demonstrated that in general practice
cohorts there was an increase in cardiovascular mortality when ACR is higher than
30 mg/g (3 mg/mmol).4 Analysis of data from the Heart Outcomes Prevention Evaluation
(HOPE) study demonstrated that any degree of albuminuria is a risk factor for cardiovascular
events in individuals with or without diabetes.
428
The lack of a threshold of albuminuria for cardiac risk was also confirmed in the
HUNT 2 Study
429
and the Losartan Intervention For Endpoint Reduction in Hypertension (LIFE) study
in patients with LVH.
430
Albuminuria and low eGFR were synergistic cardiovascular mortality risk factors in
the HUNT 2 study and using both ACR and eGFR improved cardiovascular risk stratification
at all age levels, but particularly in persons 70 years and over.
429
In the MDRD Study cohort of patients with GFR categories G3a-G4 (GFR 15–59 ml/min/1.73
m2), cystatin C level was strongly associated with all-cause and cardiovascular mortality
particularly in the elderly.
431
Analysis of data from MESA and the Cardiovascular Health Study (CHS) demonstrated
that the worst prognosis of CVD, heart failure, and progressive CKD was found in those
subjects with CKD diagnosed using the cystatin C-based equation.
432
LVH is common with CKD, is known to reflect target organ damage, and is associated
with increased cardiovascular mortality in CKD.360 It is also important to consider
the role of CKD-specific risk factors particularly in patients with more severe CKD
(GFR <30 ml/min/1.73 m2) and correct them where possible. Anemia has been thought
to have a particular role in the early development of CVD in people with CKD. Although
treatment of anemia is associated with improved well-being and greater exercise capacity,
the results of several RCTs of anemia correction have suggested that complete correction
of anemia is not advisable in people with lower GFR (<60 ml/min/1.73 m2). A meta-analysis
of 9 RCTs comparing different target Hb levels suggests an increase in mortality and
worse BP control with the higher treatment targets, independent of the GFR category.
433, 434
Those patients with LVH at baseline, a sign of target organ damage and which was present
in 47% of the Cardiovascular Risk Reduction by Early Anemia Treatment with Epoetin
Beta Trial (CREATE) cohort, had significantly worse cardiovascular outcomes when treated
to the higher targets.
435
Abnormalities of mineral metabolism with low 1,25-dihydroxyvitamin D and PTH were
found to occur early in CKD in the Study for the Evaluation of Early Kidney disease
[SEEK] study, however serum calcium and phosphate were usually normal until eGFR fell
below 40 ml/min/1.73 m2.367 An association between elevated serum phosphate and cardiovascular
mortality was demonstrated in a prospective study of people with CKD and this was
also thought to be associated with low vitamin D levels.
436
Pediatric Considerations
While the majority of data for, and evidence of, the high risk of cardiovascular morbidity
and mortality in the pediatric renal population stems from the ESRD (dialysis and
transplant) populations,
437, 438, 439, 440
there is recognition that many if not all of the factors responsible for such outcomes
will develop much earlier in this course of CKD. In fact, the current American Heart
Association's guidelines for reduction of cardiovascular risk in pediatrics stratified
the child with CKD to the highest risk category.
441
Based on data from the CKiD trial, Wilson et al.
442
demonstrated extremely high prevalence rates for four traditional cardiovascular risk
factors (hypertension, dyslipidemia, abnormal glucose metabolism, and obesity) in
a subgroup of this cohort of 586 children with iohexol GFR proven mild to moderate
CKD. In this cross-sectional analysis of 250 children who had data on all variables
of interest, median GFR of 45.2 ml/min/1.73 m2 (IQR 34.6-58.2), the prevalence of
hypertension was 46%, dyslipidemia was 44%, abnormal glucose metabolism was present
in 21%, and obesity in 15%. Hypertriglyceridemia was common, seen in 33%. In aggregate,
only 26% of the cohort had no risk factors present, 39% had at least one and 22% had
two, 11% had three and 2% had all four risk factors present. The prevalence of all
these factors remained elevated even when the population was restricted to those defined
as lean, BMI <85th. In this group, the presence of 2 or 3 cardiovascular risk factors
was still seen in 20% and 2%, respectively. Within the lean group 12% demonstrated
abnormal glucose metabolism. Finally, multivariate regression analysis of their data
demonstrated that glomerular disease as a cause of CKD and the presence of nephrotic
range proteinuria were associated with much higher odds of having cardiovascular risk
factors, 1.96 (95% CI 1.04-3.72) and 2.04 (95% CI 0.94-4.43), respectively.
It is also well-recognized that carotid intimal media thickness and pulse wave velocity
changes, as potential surrogate measures for vascular damage in children, can be increased
in in children with CKD.
443, 444
For further detailed discussions of the pathophysiology and complexity of CVD in the
pediatric CKD population, two recent review papers from Mitsnefes
445
and Shroff et al.
446
are suggested. Both papers identify and discuss the pathophysiology of both traditional
and CKD-specific cardiovascular risk factors as seen in this population.
4.1.2: We recommend that the level of care for ischemic heart disease offered to people
with CKD should not be prejudiced by their CKD. (1A)
4.1.3: We suggest that adults with CKD at risk for atherosclerotic events be offered
treatment with antiplatelet agents unless there is an increased bleeding risk that
needs to be balanced against the possible cardiovascular benefits. (2B)
RATIONALE
These statements are worded in this way to reflect the growing body of evidence that
suggests that treatment of traditional risk factors in CKD patients is of benefit,
and that attempts to modify or preclude usual investigations and therapies are not
necessary in CKD patients. Inasmuch as CKD populations have been understudied, but
likely inadvertently included in larger population studies, the rationale for risk
factor modification is similar to that in the general population. Modification of
risk factors includes:
Smoking cessation
Exercise
Weight reduction to optimal targets
Lipid modification recognizing that the risk reduction associated with statin therapy
in adults with CKD is relatively constant across a broad range of baseline low-density
lipoprotein cholesterol (LDL-C) levels
Optimal diabetes control HbA1C <7% (53 mmol/mol)
Optimal BP control to <140/90 mm Hg or <130/80 mm Hg in those with CKD and depending
on the degree of proteinuria (see Recommendations 3.1.4 and 3.1.5)
Aspirin is indicated for secondary prevention but not primary prevention
Correction of anemia to individualized targets (see KDIGO Clinical Practice Guideline
for Anemia in CKD
11)
Evidence Base
The maintenance of cardiovascular health ultimately impacts ‘kidney health' given
the links between cardiac function, renal perfusion, atherosclerosis, and glomerulosclerois.
The details of this complex biology are beyond the scope of these guidelines, but
can be found in key references below.
There is evidence in the general population that smoking cessation is associated with
a reduction in cardiovascular risk. In CKD there is evidence that smoking is associated
with progression of renal disease, though no specific data exist that support cessation
of smoking and delay of progression.
447
People with CKD have reduced levels of muscle strength and aerobic activity. In those
not on dialysis, exercise training is shown to improve functional capacity and BP.
448
Weight loss in obese CKD patients can reduce rate of decline, proteinuria, and BP.354
The Study of Heart and Renal Protection (SHARP)
449
is the largest RCT in people with CKD to date. The results demonstrate that a lipid-lowering
strategy which included fixed dose simvastatin and ezetimibe resulted in a 17% reduction
in atherosclerotic events, as compared to placebo. The cohort enrolled included those
with eGFR under 60 ml/min/1.73 m2, age greater than 40 years, and was an international
study including more than 9000 subjects from around the globe. The lipid-lowering
strategy was effective and safe. Lipid targets have not been established specifically
for people with CKD. Treatment strategies should be implemented in accordance with
current recommendations for high-risk populations.
At the time of this writing, the KDIGO Clinical Practice Guideline for Lipid Management
in Chronic Kidney Disease was under preparation for public review. In brief, the key
aspects of the draft recommendations include treating those at high risk for atherosclerotic
disease with lipid-lowering therapies, regardless of LDL levels, in those 50 years
of age and above. Since this Guideline has not yet been finalized, interested readers
should refer to the final document when it is formally released in 2013.
The benefits of aspirin in people with CKD and hypertension was demonstrated in a
post hoc analysis of the Hypertension Optimal Treatment (HOT) trial.
450
Jardine et al. reported that among every 1000 persons with eGFR <45 ml/min/1.73 m2
treated for 3.8 years, 76 major cardiovascular events and 54 all-cause deaths will
be prevented while 27 excess major bleeds will occur. They concluded that an increased
risk of major bleeding appears to be outweighed by the substantial benefits. Clopidogrel
is used as an alternative to aspirin but CKD has been shown to be associated with
an increase in platelet reactivity and there is resistance to clopidogrel in people
with CKD, diabetes, and CVD.
451
The Clopidogrel for Reduction of Events During Observation (CREDO) trial concluded
that clopidogrel in mild or moderate CKD may not have the same beneficial effect as
it does in people without CKD. Subjects with normal renal function who received 1
year of clopidogrel had a marked reduction in death, MI, or stroke compared with those
who received placebo (10.4% versus 4.4%, P <0.001), whereas those with mild and moderate
CKD did not have a significant difference in outcomes with clopidogrel therapy versus
placebo (mild: 12.8% versus 10.3%, P = 0.30; moderate: 13.1% versus 17.8%, P = 0.24).
Clopidogrel use was associated with an increased RR of major or minor bleeding, but
this increased risk was not different based on renal function.
452
International Relevance
Although it is clear and mentioned in many guidelines that CKD is associated with
an increase in CVD risk, it is not included in many assessment tools and there is
a deficiency in ethnicity and regional specific tools.
The assessment tool from US Third Report of the National Cholesterol Education Program
(NCEP III) does not include CKD. The Fourth Joint Task Force of the European Society
of Cardiology suggest that both an eGFR of less than 60 ml/min/1.73 m2 and albuminuria
increase risk but do not quantify the risk or include CKD in their Systematic COronary
Risk Evaluation (SCORE) assessment tool.
453
In the UK, the QRISK®2 online tool includes CKD as ‘yes' or ‘no' and therefore does
not allow for level of eGFR or proteinuria.
454
The Joint British Societies Guidelines regards proteinuria as a sign of target organ
damage, which conveys a risk of at least 20% in 10 years.
455
A study of cardiovascular risk estimation in Chinese adults in the USA–People's Republic
of China Collaborative Study of Cardiovascular and Cardiopulmonary Epidemiology (USA-PRC
Study) did not include CKD as a risk factor at all.
456
It is also important to remember that in this population stroke was the predominant
CVD. A review of 25 risk assessment tools identified only 2 derived from an Asian
population.
457
However, links between GFR categories and CVD events are evident in the Asian population,
for example in the Japanese Gonryo study.
458
A prospective study in a general Japanese population demonstrated links between lower
GFR, high creatinine levels, and proteinuria with cardiovascular mortality particularly
stroke.
459
In Chinese patients who were at least 50 years old and either had existing CVD or
were at high risk, 34% had an eGFR <60 ml/min/1.73 m2 and eGFR <45 ml/min/1.73 m2
was found to be an independent predictor of all-cause and cardiovascular death.
460
In a population-based study of 2353 people aged over 40 years in Beijing, an eGFR
<90 ml/min/1.73 m2 was associated with increased CVD risk, and for each CKD category
stroke was more prevalent than MI.
461
In India the increasing prevalence of type 2 diabetes is driving an increase in CKD
and both are associated with increased cardiovascular risk. In CKD patients attending
one clinic in North India, 28% had diabetes, 27% were overweight, and 92% had hypertension.
Metabolic syndrome as defined by the International Diabetes Federation (IDF) 2007
guidelines was present in 39% and more frequent in women.
462
Implications for Clinical Practice and Public Policy
A full CVD risk assessment in a person with CKD should include an estimate of GFR
and a quantitative assessment of albuminuria. Cystatin C may be helpful at risk stratifying
those at intermediate risk.
The key aspect of these recommendations is to ensure that people with CKD are not
deprived of treatment strategies known to be effective in general populations.
Areas of Controversy, Confusion, or Non-consensus
Many of the traditional cardiovascular risk assessment tools do not adjust for the
presence of early CKD. Any tool should also adjust for level of eGFR because of increasing
risk associated with lower levels of kidney function. The Framingham risk equation
underestimates true events in a CKD population, but no validated different tools exist
at present which better quantify cardiovascular or mortality risk in CKD populations.
Proteinuria is considered to be a sign of target organ damage and thus associated
with high cardiovascular risk. However, assessment of risk should also include the
presence of albuminuria. Few studies exist which demonstrate that targeting lowering
of proteinuria results in reduced cardiovascular risk, though a number of studies
which have targeted interruption of the RAAS (and in which lowering of urine protein
has also been shown) have demonstrated a benefit.
An appropriate risk assessment tool should be available to assess risk in people with
CKD. Addition of the presence of CKD to conventional risk factors in the Reykjavik
population increased discrimination but did not increase risk to the same degree as
smoking or diabetes.
463
In the Framingham Heart Study, GFR category 3b (GFR 30–44 ml/min/1.73 m2) was associated
with CVD but not equivalent to previous CVD.
464
Lifestyle modification has not been studied in large trials in people with CKD but
smoking cessation, achievement of optimum weight, regular exercise, and salt restriction
should be seen as reasonable aims.
Pediatric Considerations
It is presently unclear whether children with CKD and elevated lipid levels will benefit
from lipid-lowering strategies with no prospective trials ongoing in this area. It
is hoped that data from both CKiD55 and the 4C trials78 may be able to address these
shortcomings.
However, given that these children demonstrably have elevated risk of cardiovascular
and atherosclerotic disease (see Recommendation 4.1.1), that there are data supporting
the use of a number of statins in the pediatric population,
465
and that the adult evidence for benefit of statins in a subset of the SHARP trial
449
with a GFR <60 ml/min/1.73 m2 exist, it would seem reasonable to consider the use
of such drugs in children with CKD and elevated lipids.
Recommendations from 2006
441
for use of dietary modifications, followed by statins in children older than 8 years
of age with persistent elevation in LDL-C levels, have been endorsed by the American
Academy of Pediatrics. However, at the time of publication of these guidelines, the
KDIGO Dyslipidemia Work Group has suggested that statin therapy may not be appropriate.
Age- and renal function-adjusted doses of such drugs should be carefully considered
prior to any therapy being initiated (see Recommendation 4.4.1).
Regarding diabetes control, treatment in keeping with national and international diabetes
recommendations is prudent. Note is made that there are specific caveats with respect
to drugs and side effects that are important (see Recommendations 3.1.15-3.1.18).
There is no literature in the area of antiplatelet agents in atherosclerotic disease
in the subset of children with CKD and the suggestion to offer antiplatelet agents
does not apply to pediatric practice.
4.1.4: We suggest that the level of care for heart failure offered to people with
CKD should be the same as is offered to those without CKD. (2A)
4.1.5: In people with CKD and heart failure, any escalation in therapy and/or clinical
deterioration should prompt monitoring of eGFR and serum potassium concentration.
(Not Graded)
RATIONALE
This statement is worded this way to make clinicians aware that people with CKD and
heart failure should receive at least equal benefit from heart failure therapy as
those without CKD, but that heart failure therapies (in particular increase of RAAS
blockade and diuretic therapy) may lead to significant changes in GFR and serum potassium
concentrations. This does not imply that such therapy should be avoided but only that
clinicians are cognizant of this possibility, monitor it, and understand it in the
context of individual risks and benefits.
Heart failure is a complex clinical syndrome which can be caused by any structural
or functional cardiac disorder that impairs the pump function of the heart and has
a high mortality. Within the general population, the commonest causes of heart failure
are ischemic heart disease causing left ventricular systolic dysfunction, and hypertensive
heart disease with left ventricular hypertrophy and diastolic dysfunction. Thus people
with CKD are at increased risk of both. Evidence-based medication for heart failure
impacts on GFR and prescribers need to be alert to this. Changes in GFR are often
transient and the downward trajectory GFR is not sustained if all other aspects of
care and clinical status are stable. Monitoring and close follow-up are required.
Evidence Base
In an outpatient setting, normal GFR is unusual in heart failure. A study from the
Alberta Heart Function Clinic found that in people with an ejection fraction of ≤35%,
GFR categories G3a and G3b (GFR 30–59 ml/min/1.73 m2) were present in 40% and 16%
of people were GFR categories G4 or G5 (GFR<30 ml/min/1.73 m2).
466
This study demonstrated a 1% increase in mortality associated with each 1 ml/min fall
in CrCl demonstrating the clear link between kidney function and outcome for people
with CKD and heart failure. Meta-analysis of 16 studies and over 80,000 people with
heart failure by Smith et al. showed that renal impairment was present in 63% and
was associated with increased mortality across the range of kidney function, with
an adjusted HR of 1.56 (95% CI 1.53-1.60).
467
In the Candesartan in Heart Failure-Assessment of Reduction in Mortality and Morbidity
(CHARM) Program with candesartan in heart failure, 36% of subjects had eGFR <60 ml/min/1.73
m2 which was associated with increased mortality and hospital admission regardless
of ejection fraction.
468
The Digitalis Intervention Group (DIG) identified a GFR of 50 ml/min as a risk threshold
for increased mortality risk in heart failure.
469
Smith et al. demonstrated that 1-year mortality after admission for heart failure
was 29% in people with normal kidney function but 52% in those with moderate to severe
renal impairment.
470
Anand et al. also demonstrated an independent association between the presence of
positive reagent strip proteinuria present in 8% of subjects and increased mortality
in heart failure.
471
In the Alberta Provincial Project for Outcome Assessment in Coronary Heart Disease
(APPROACH) study of people with heart failure related to coronary artery disease (CAD),
39% had CrCl <60 ml/min (<1 ml/s). This study also demonstrated that the level of
kidney function was a better indicator of poor outcome than cardiac anatomy.
472
In people admitted to hospital with heart failure, increasing SCr concentration is
associated with a longer hospital stay and worse outcome.
473
LVH is present in 40% of people with CKD,
474
progressively increasing as kidney function declines.
475
Ha et al. demonstrated LVH in 87% of people with predialysis CKD.
476
Parfrey et al. in a study of 432 people who were predialysis, found that only 16%
had a normal echocardiogram while 41% had concentric LVH, 16% systolic dysfunction,
and 28% left ventricular dilatation.
477
LVH is related to increased work load caused by arterial stiffness and hypertension,
and at lower levels of GFR, volume overload. Alterations in electrolyte balance, anemia,
bone metabolism, uremia, oxidative stress, inflammation, and other inflammatory mechanisms
all play a role.
478
In the majority of people with CKD, the myocardium develops concentric hypertrophy
476
with interstitial fibrosis and CKD-associated cardiomyopathy, leading to left ventricular
stiffness.
474
This results in diastolic heart failure; people with CKD have higher mortality from
diastolic heart failure than systolic heart failure.
479
Diastolic heart failure is more common in people with CKD who are older females with
hypertension and/or diabetes.
480
Left ventricular hypertrophy is particularly associated with anemia and a fall in
Hb to less than 12.8 g/dl (128 g/l) is associated with left ventricular growth in
people with early CKD.
481
Lower Hb levels are associated with increased risk of death and hospitalization in
heart failure.
482
Other cardiovascular conditions, for example MI and atrial fibrillation, are more
common in CKD and may exacerbate heart failure.
It is important to note that in a study of an older population (age >64 years), heart
failure was an independent predictor of rapid kidney function decline.
483
Standard therapy for the management of systolic heart failure includes an ACE-I or
ARB and a beta-blocker licensed for heart failure such as bisoprolol or carvedilol.
Meta-analysis of 25 trials of RAAS blockade (ACE-I or ARB) in people with CKD or proteinuria,
demonstrated a reduction in the risk for heart failure and reduction in cardiovascular
outcomes.
484
Subgroup analysis from the Valsartan Heart Failure Trial (Val-HeFT) demonstrated the
beneficial effect of valsartan in people with CKD and that valsartan reduced the eGFR
by the same amount in people with and without CKD.
471
Candesartan was also clinically effective in the CHARM study regardless of underlying
kidney function
468
while irbesartan was associated with a 23% reduction in hospital admissions related
to heart failure in people with nephropathy due to type 2 diabetes.
485
Analysis of data on bisoprolol from the Cardiac Insufficiency Bisoprolol Study II
(CIBIS II)
486
identified 32% of subjects with eGFR <60 ml/min/1.73 m2. These subjects were more
likely to die or be admitted to hospital than those with eGFR≥60 ml/min/1.73 m2. However,
they showed a similar benefit from bisoprolol treatment. A post hoc analysis of data
on carvedilol from the Carvedilol Post-Infarct Survival Control in Left Ventricular
Dysfunction (CAPRICORN) and Carvedilol Prospective Randomized Cumulative Survival
(COPERNICUS) trials identified 60.8% of participants with eGFR <60 ml/min/1.73 m2.
487
Carvedilol was well tolerated by subjects with CKD, and decreased all-cause, cardiovascular,
and heart failure mortality. However, doubts were raised over the benefits when eGFR
<45 ml/min/1.73 m2.
Aldosterone antagonists are included in standard heart failure management, usually
in addition to an ACE-I or ARB, leading to concerns relating to risk of hyperkalemia,
particularly in people with lower GFRs. In a study of spironolactone therapy in severe
heart failure, Pitt et al.
488
included subjects with SCr concentrations up to 2.5 mg/dl (221 μmol/l). There was
a 30% reduction in mortality with spironolactone and the incidence of hyperkalemia
was low. However, a review of real-world practice demonstrated a significant increase
in hyperkalemia with the combination of RAAS blockade and an aldosterone antagonist,
highlighting the need for close monitoring with the introduction of combination therapy.
489
In the APPROACH study on the specialist management of heart failure, subjects with
CAD and CKD, despite being at high risk, were less likely to be taking ACE-Is, beta-blockers,
statin, and aspirin. This study demonstrated that those subjects taking beta blockers
had lower mortality, however ACE-Is were only of benefit when eGFR was >60 ml/min/1.73
m2.
472
International Relevance
The treatment of heart failure is similar around the globe. The ability to closely
monitor kidney function or to offer conventional therapies may differ, however. Irrespective,
this statement should hold true internationally. The National Heart Care Project of
community subjects admitted to hospital with heart failure studied the differences
between people of black and white ethnic groups. People with worse renal function
were more likely to be black, older and female, and black people had a greater prevalence
of hypertension and diabetes but less ischemic heart disease. Black people had a lower
risk of mortality at every level of creatinine, for every 0.5 mg/dl (44.2 μmol/l)
increase in creatinine, 1-year death risk increased 10% in black people and 15% in
white people.
470
Implications for Clinical Practice and Public Policy
Cardiorenal syndrome, an impairment of kidney function in the presence of heart disease,
is a marker of worse prognosis. In addition to GFR, the presence and severity of albuminuria
and anemia aid in prognostication and management.
490
Where the link between LVH and heart failure in CKD is so strong and the outcome poor
for people with LVH, efforts should be made to optimize care particularly BP. There
is a lack of robust evidence for the optimal medical management of heart failure specifically
in people with CKD; however, the prevalence of people with early CKD is so high in
heart failure trials that we can apply standard treatment to these people. For those
people with eGFR <45 ml/min/1.73 m2, the situation is less clear.
487
Areas of Controversy, Confusion, or Non-consensus
There are difficulties in making the diagnosis of heart failure in people with CKD,
especially diastolic heart failure. Biomarkers such as B-type natriuretic peptide
(BNP) may be abnormally elevated in CKD, but may or may not be as responsive to treatment
or diagnostically accurate in people with CKD (see Chapter 4.2). There is little evidence
to guide the management of left ventricular systolic dysfunction in people with CKD
and evidence for the management of diastolic heart failure is generally lacking. However,
there is evidence that medications to improve the outcome of people with heart failure
are underused in people with CKD.
Appropriate management of the metabolic complications of kidney disease which can
exacerbate heart failure needs to be clarified. Clinical studies are required on the
role of device therapy in people with CKD (e.g., pacemakers, defibrillators, etc.).
4.2 CAVEATS WHEN INTERPRETING TESTS FOR CVD IN PEOPLE WITH CKD
BNP/N-terminal-proBNP (NT-proBNP)
As GFR diminishes, the prevalence and severity of cardiovascular abnormalities increase
and are accepted as major causes of morbidity and mortality in people with CKD. The
presence of congestive heart failure (CHF) in people with CKD conveys a worse prognosis
than either condition alone. Therefore, early diagnosis and aggressive management
of CHF are both highly desirable. A number of cardiac biomarkers (i.e., BNPs and cardiac
troponins) appear to have both increasing clinical importance and increasing facility
for detection and stratification of CHF. The stimulus for secretion of these biomarkers
is the hemodynamic load (i.e., myocardial stretch) and their secretion is associated
with the severity of CHF and the degree of left ventricular dysfunction. They are
thus useful markers for diagnosis, management, and prognosis in people with normal
renal function. However, when the eGFR is less than 60 ml/min/1.73 m2, the accuracy
of plasma BNP and NT-proBNP levels for detection and stratification of CHF becomes
unreliable and the degree of responsiveness to treatment is not known.
4.2.1: In people with GFR <60 ml/min/1.73 m2 (GFR categories G3a-G5), we recommend
that serum concentrations of BNP/NT-proBNP be interpreted with caution and in relation
to GFR with respect to diagnosis of heart failure and assessment of volume status.
(1B)
RATIONALE
This statement is intended to remind clinicians that although the prevalence of fluid
overload and heart failure increases with lower GFR categories, BNP becomes less reliable
as a predictor of fluid overload and heart failure at lower GFRs. While it is associated
with worse outcomes, cutoff values indicative of heart failure in general populations
may or may not be appropriate and changes in values with treatment may or may not
have the same meaning.
Evidence Base
Natriuretic peptides belong to a family of circulatory peptide hormones from either
myocardial cell origin (atrial natriuretic peptide [ANP] and BNP) or endothelial cell
origin (C-type).
491, 492
They play an important role in regulating BP and body fluid volume by their natriuretic
and diuretic actions, arterial dilatation, and inhibition of the RAAS.
493
Natriuretic peptides for the diagnosis of CHF have been a diagnostic breakthrough
in cardiology where determination of their concentration in serum can help identify
patients with left ventricular systolic dysfunction.
494, 495
Concentrations of natriuretic peptides increase in people with CHF and other CVDs
as a consequence of pressure and volume overload.
496, 497
Among people with progressive LVH, concentrations of NT-proBNP are typically elevated
in proportion to the degree of left ventricular mass increase.
Overall, BNP levels provide a better index of left ventricular mass and load than
do ANP levels and BNP has emerged as a superior biomarker to ANP for clinical applications
involving CHF and left ventricular dysfunction.
491, 498
NT-proBNP may have analytical advantages over BNP because of greater stability due
to a longer half-life. The cause of an elevated BNP level is multifactorial in origin
and may reflect cardiac dysfunction, and/or changes in renal function.
499
Since renal dysfunction intrinsically was shown to affect BNP levels in some studies,
the diagnostic value of BNP levels in the presence of CKD has been questioned.
500, 501
Earlier research hypothesized that measurement of natriuretic peptides could be of
use in monitoring excess fluid volume and dry weight in people requiring dialysis,
502, 503
but this remains unproven and the significance of plasma natriuretic peptides and
their clinical role in dialysis remains unclear.
491
Although several studies revealed higher BNP levels as GFR declined, outcomes of other
studies suggest that the relationship between renal function and BNP levels may be
most strongly dependent on cardiac and volume-related factors.
504, 505, 506
To test this hypothesis, Tagore et al. studied BNP levels in a cohort of 143 clinically
euvolemic subjects with CKD in whom absence of heart disease was clinically validated
and found that plasma BNP levels were independent of GFR.
504
Suresh and Farrington
507
studied people on dialysis and concluded that BNP levels were predictive of presence
of left ventricular dysfunction, cardiac events, and survival in the presence of ESRD,
suggesting that BNP levels may be informative across the full range of renal function
and even in its absence. As previously established by several studies, the level of
NT-proBNP is a strong prognostic marker in both the general population
508
and in various disease states, e.g., acute and CHF,
509, 510
coronary heart disease,
511, 512
and hypertension.
513
International Relevance
The issue of how to differentially interpret BNP in the presence or absence of CHF
and within the context of declining renal function has received considerable attention
but remains incompletely resolved from a clinical perspective. Changes in either renal
or cardiac function could potentially impact levels of natriuretic peptides and incomplete
knowledge of an individual's clinical status relative to the degree of cardiac pathology
or renal dysfunction could confound interpretation of BNP levels. Given the cost of
biomarkers assays and the uncertainties related to interpretation of serum levels
of BNP/NT-proBNP in people with CKD, competing priorities for health-care resource
may dictate reliance on clinical evaluation and local practices of care in some areas
of the world. Until better data are available, clinicians are asked to use this laboratory
measurement with caution as a diagnostic test for heart failure in those with CKD.
Implications for Clinical Practice and Public Policy
Since people with ESRD with elevated levels of BNP have a higher RR of death, this
biomarker could be used as an individual risk marker with the aim of identifying those
at high risk. Clinical correlation is of utmost importance to ensure accurate diagnosis
and appropriate therapy.
Studies to understand the variability of BNP by categories of eGFR and urine ACR,
within and between individuals over time, and the response to therapy should be undertaken
so as to better inform clinicians.
Areas of Controversy, Confusion, or Non-consensus
The relationship between renal function and BNP, the role of natriuretic peptides
as markers of fluid overload and as predictors of mortality, and the utility of serial
monitoring of BNP/NT-proBNP levels in people with CKD, and responsiveness to therapy
remains unknown.
Clarification of issues and key points
Natriuretic peptides may be surrogate markers for prediction of mortality in people
with CKD reflecting the association of mortality with left ventricular dysfunction
and LVH. Before routine measurements of these biomarkers are recommended, their utility
in guiding or changing clinical practice should be assessed.
Pediatric Considerations
There is robust literature in both children and neonates regarding both the normative
values of BNP
514
and its value in predicting outcomes of pediatric heart failure.
515
To date no studies in children specifically address the issue of reduced GFR in this
population. However, application of similar principles would likely apply.
Troponins
In recent years newer markers of myocardial injury have been introduced into clinical
practice.
516, 517
Among these, cardiac troponins have proven to be specific markers of myocardial damage.
518, 519
Cardiac troponins have been considered the gold standard biochemical test for diagnosis
of myocardial damage because they have nearly absolute myocardial tissue specificity,
as well as high sensitivity, thereby reflecting even microscopic zones of myocardial
necrosis.
517, 520
Furthermore, increasing evidence indicates that abnormal troponin measurements identify
a subgroup of patients who have an increased risk of major cardiac events
519, 521, 522, 523, 524, 525
and that measurement of cardiac troponin T (cTnT) prior to commencing renal replacement
is a significant independent predictor of survival.
526
However, these markers are renally excreted and are often found to be elevated in
people with CKD irrespective of specific symptomatology. Thus, there is uncertainty
as to the clinical importance of their different magnitudes of elevation in CKD. There
are no data on ‘expected' values of these markers in the context of CKD, or at different
categories of GFR and albuminuria.
4.2.2: In people with GFR <60 ml/min/1.73 m2 (GFR categories G3a-G5), we recommend
that serum concentrations of troponin be interpreted with caution with respect to
diagnosis of acute coronary syndrome. (1B)
RATIONALE
This statement is intended to remind clinicians that isolated elevated serum concentrations
of troponin (in the absence of symptoms or electrocardiographic changes) may not necessarily
portend a diagnosis of ACS at lower GFR categories. While there are data linking troponin
elevations with poor outcomes, the specific implications for an individual in clinical
practice are not clear. Hence, this statement simply cautions the clinician with respect
to interpretation but does not suggest that the laboratory values are not of significance.
In a patient with clinical symptoms suggestive for ischemic heart disease, an elevated
cardiac troponin I (cTnI) is suggestive for ACS.
Evidence Base
cTnT and cTnI are low-molecular-weight proteins that form part of the troponin complex
and are integral components of the myofibrillar contractile apparatus of the heart.
527
Loss of integrity of cardiac myocyte membranes causes release of cardiac troponins
into the circulation, which can be detected by highly sensitive assays developed for
cTnT and cTnI to diagnose ACS.
521
A potent stimulus for the release or production of natriuretic peptides is mechanical
stretching,
528
whereas plasma cTnT specifically and sensitively reflects myocardial injury and is
considered to be an index of irreversible myocardial change.
521
In addition, cTnT levels can predict multi-vessel CAD in people requiring dialysis.
529
However, the significance of screening people with CKD for CAD using these cardiac
biomarkers is questionable
530
since increases in serum cardiac troponin concentration can occur in the absence of
an ACS.
531
Most studies
517, 532, 533, 534, 535, 536, 537, 538, 539, 540, 541, 542, 543, 544, 545
have focused on people with ESRD where increases in serum cTnT concentrations have
been observed in 20%–90% of subjects
540
but generally much lower when cTnI was measured. Information about cardiac troponins
and their relationship with comorbidity is sparse in people with CKD who are not receiving
dialysis treatment.
526, 527, 546, 547
Increased cardiac troponin concentrations may occur early in CKD, including GFR categories
30-59 ml/min/1.73 m2, and more commonly as CKD advances.
Powerful prognostic data emerging from the dialysis population demonstrate that increased
concentrations of cardiac troponins reflect either subclinical myocardial damage caused
by silent ischemia or myocardial remodeling in the development of LVH.
539, 542
Increased levels of cTnT measured with the highly sensitive assay, well below the
detection range of standard assays, have been shown to be associated with cardiac
structural abnormalities including LVH (both left ventricular wall thickening and
dilation) and left ventricular systolic dysfunction,
548
and to have a graded association with all-cause and CVD mortality independent of traditional
risk factors, renal function, and levels of other biomarkers such as high sensitivity
CRP and NT-proBNP.
These findings suggest that chronic elevation of troponin levels may be mediated to
a greater extent by indices of heart failure (such as higher left ventricular mass,
left ventricular dysfunction, or increased NT-proBNP levels) than indices of atherosclerosis
or ischemia.
548
International Relevance
The costs of these tests, the differing sensitivity of the assays available, the paucity
of data in people with CKD, and the poor correlation of serum troponin levels with
ACS in people with CKD should be weighed against competing local health-care priorities
and resource before their introduction into routine clinical practice. Emphasis should
be placed on the clinical context and local standard practices of care.
Implications for Clinical Practice and Public Policy
Troponin elevation may result from repeated episodes of clinically silent MI,
522, 549
although CAD frequently underlies troponin elevation and increased cTnT is also frequently
encountered in patients at low risk for CAD. Troponin elevations are as frequent in
subjects without CAD as in those with known CAD
517, 550, 551, 552
and in people with CKD, serial increasing cTn levels may be of greater utility for
diagnosis of ACS. In the context of symptoms and signs of ACS, elevations of troponins
should not be ignored nor incorrectly attributed to simply ‘CKD.' Thus the importance
of ordering the test for specific indications for those with CKD should be emphasized.
Areas of Controversy, Confusion, or Non-consensus
Information about cardiac troponins and their relationship with comorbidity is sparse
in people with CKD who are not receiving dialysis treatment. Although elevated values
for TnT are common in people with low GFR in the absence of ACS or CHF, the prevelance
of elevated TnI in the absence of ACS or CHF in people with low GFR may be only 0.4-6%
depending on the cutoff value chosen.
532
Clarification of Issues and Key Points
As pointed out by Roberts et al.,
553
although management strategies in people without CKD based on these biomarkers are
quite clear, in people with CKD it is not the case despite the strong data on the
prognostic implications. These biochemical markers can only be useful if they are
interpreted in the context of the clinical history and examination with consideration
given as to what management is appropriate for each individual, and if a cardiovascular
therapy that is effective in this population is identified or developed.
553
The importance of understanding and contextualizing elevations in troponin values
in CKD populations cannot be overstated: under-treatment in symptomatic individuals
should be avoided, and better understanding of treatment strategies in those asymptomatic
individuals is the focus of ongoing study.
Pediatric Considerations
While normative values exist for troponin in children,
554
its utility as a general screening marker for myocardial ischemia in pediatrics has
recently been called into question,
555
and no data exist relating reduced renal function and troponin levels in children.
Non-invasive testing
Most studies of cardiac evaluation in people with CKD have involved candidates for
kidney transplantation in order to identify existing cardiac conditions amenable to
risk modification and to exclude people with short expected survival due to cardiac
morbidity. Among the tests for cardiovascular assessment, coronary angiography is
considered invasive, expensive, and associated with risks such as contrast-induced
nephropathy and cholesterol embolism. When trying to detect CAD or predict future
cardiovascular events in people with CKD, nuclear imaging modalities using thallium
tracers have shown conflicting results, especially in the context of small vessel
disease. Clinicians should be aware of the limitations of radionuclear testing as
a screening tool for significant CAD given the limitations of the test. Exercise-dependent
techniques are limited by poor exercise tolerance of many subjects and people with
CKD frequently have cardiac evaluation performed with pharmacological agents.
556, 557
4.2.3: We recommend that people with CKD presenting with chest pain should be investigated
for underlying cardiac disease and other disorders according to the same local practice
for people without CKD (and subsequent treatment should be initiated similarly). (1B)
4.2.4: We suggest that clinicians are familiar with the limitations of non-invasive
cardiac tests (e.g., exercise electrocardiography [ECG], nuclear imaging, echocardiography,
etc.) in adults with CKD and interpret the results accordingly. (2B)
RATIONALE
These statements are worded in this way firstly because although people with CKD are
at increased risk from cardiovascular events, observational study has shown that they
are frequently discriminated against in terms of investigation and management of such
events. Secondly, non-invasive cardiac tests have known additional limitations in
people with CKD and such limitations and their consequences need to be clearly understood
by health professionals caring for people with CKD.
Evidence Base
Noninvasive stress testing appears to be the most common first approach to cardiac
evaluation of asymptomatic patients, prompted by diabetes, age and risk-factor burden.
These tests include myocardial perfusion studies, stress echocardiography, and most
recently, cardiac computed tomographic angiography. However, people with CKD are underrepresented
in studies evaluating the diagnostic sensitivities and specificities of noninvasive
tests. Exercise ECG is limited by lack of specificity of the ST-segment response and
by inability of many people with CKD to exercise to a diagnostic workload.
These tests have imperfect sensitivity and specificity in people with CKD or, in the
case of tomographic angiography, evaluation in this population is as yet unpublished.
556
CAD has been detected in a high proportion of people on long–term dialysis therapy.
Studies describing associations of angiographic coronary stenosis with subsequent
clinical events in people with ESRD, including those undergoing transplant evaluations,
have reached inconsistent conclusions.
558, 559, 560, 561, 562
Abnormalities on myocardial perfusion studies correlate well with the presence of
CAD in the general population, with mean weighted sensitivity of 88% and mean weighted
specificity of 74%.
563
The performance of myocardial perfusion studies in identifying CAD in people with
CKD is more variable, with reported sensitivities and specificities in people with
ESRD ranging from 37% to 90% and 40% to 90%, respectively.
564, 565, 566, 567
Nonetheless, results from myocardial perfusion studies have prognostic value for cardiac
events and mortality.
568, 569
In a meta-analysis of 12 studies involving thallium-201 scintigraphy and dobutamine
stress echocardiography, people with ESRD with inducible ischemia had approximately
6 times the risk of MI and 4 times the risk of cardiac death as those without inducible
defects.
570
Moreover, subjects with fixed defects also had nearly 5 times the risk of cardiac
death.
The prognostic value of myocardial perfusion studies has been shown with other perfusion
tracers. For example, in a study of 126 people with ESRD who underwent technetium-99m
myocardial perfusion studies as part of their pre-transplant assessment, the presence
of a reversible defect was associated with 3 times the risk of post-transplant cardiac
events (HR 3.1; 95% CI 1.1-18.2) and nearly twice the risk of death (HR 1.92; 95%
CI 1.1-4.4) compared with normal test results.
571
De Lima et al.
572
prospectively studied 126 renal transplant candidates clinically classified as moderate
(age >50 years) or high (diabetes, extra cardiac vascular disease, or known CAD) coronary
risk with myocardial perfusion studies, dobutamine stress echocardiography, and coronary
angiography. Significant CAD, defined as greater than 70% stenosis in 1 or more major
epicardial artery on angiography was found in 42% of the sample. After a median follow-up
of 46 months, clinical risk stratification and coronary angiography predicted major
cardiac events, but myocardial perfusion studies and dobutamine stress echocardiography
did not.
Newer imaging techniques are under current evaluation but a review here is beyond
the scope of this guideline; these techniques include: stress myocardial perfusion
single photon emission computed tomography (SPECT);
557, 573, 574, 575, 576
18F-BMS, a novel fluorine-based myocardial perfusion tracer;
557, 577
and gated imaging using radio-labelled ammonia (13NH3).
International Relevance
People with CKD who are being investigated prior to transplantation should be assessed
according to the national and international cardiology guidelines and taking into
account local practice and availability of resource.
Implications for Clinical Practice and Public Policy
There is still a need for a safe, noninvasive, diagnostic modality that will allow
stratification of cardiovascular risk among people with CKD. With the development
of new imaging markers, it becomes difficult for clinicians to choose the imaging
tests that can best aid clinical decisions for a given patient. Imaging methods should
provide accuracy and precision at acceptable cost. Until advanced imaging techniques
(e.g., cardiac MRI, positron emission tomography, cardiac computed tomography) are
standardized, clinicians should rely on the most reliable and familiar techniques
in their environment.
578
Areas of Controversy, Confusion, or Non-consensus
Clinical and cost-effectiveness could be improved; cardiac evaluation in people with
CKD should be based on data from clinical trials.
Clarification of Issues and Key Points
If cardiac evaluation is at present predominantly used for risk stratification and
prognosis, in the future it may guide optimal therapy and monitor clinical progress.
557
4.3 CKD AND PERIPHERAL ARTERIAL DISEASE
There is a strong link between CKD and peripheral arterial disease (PAD). Symptoms
of PAD may only be present in a minority of people who have clinical evidence of PAD.
579
It is therefore important to measure ankle-brachial index and perform regular systematic
assessment of the lower limbs of people at high risk of PAD to identify bruits, loss
of pulses, cool pale extremities, delay in venous filling, and skin ulceration.
580
4.3.1: We recommend that adults with CKD be regularly examined for signs of peripheral
arterial disease and be considered for usual approaches to therapy. (1B)
4.3.2: We suggest that adults with CKD and diabetes are offered regular podiatric
assessment. (2A)
RATIONALE
These statements are worded in this way in order to highlight the increased risk of
PAD in people with CKD, particularly those with CKD and diabetes.
Evidence Base
The prevalence of PAD is common in people with CKD and it increases with lower levels
of GFR (Table 31). PAD can be attributed to the greater prevalence of traditional
risk factors such as diabetes, hypertension, dyslipidemia, advanced age, and renal-specific
factors.
581, 582, 583, 584, 585, 586, 587, 588
Diabetes is the leading cause of CKD globally and the risk of diabetic foot ulceration
and lower extremity amputation in subjects with diabetes increases dramatically with
lower GFR. In a cohort of 90,617 individuals with diabetes over a median observation
time of 2.4 years, the HR for amputation increased from 2.08 (95% CI 1.68–2.58) for
those with GFR 30-59 ml/min/1.73 m2 to 7.71 (95% CI 5.29–11.26) for those with GFR
<30 ml/min/1.73 m2 compared to reference subjects with GFR >60 ml/min/1.73 m2.
589
Screening for PAD is recommended for adults in the general population based on age
and number of risk factors.
590
Although an ankle-brachial index of <0.9 is generally considered evidence of PAD,
the utility of this measure is unproven in CKD because of the greater prevalence of
vessel calcification. Evidence-based medical therapies for PAD in people with CKD
are similarly lacking. Smoking cessation is mandatory. Aspirin may be beneficial for
prevention of cardiovascular events. Clopidogrel has not been studied in people with
CKD and PAD. No RCTs evaluating percutaneous versus surgical revascularization techniques
have been conducted in people with CKD and PAD but outcome studies all suggest that
CKD confers an increased risk of adverse outcome, regardless of the revascularization
technique employed.
591, 592
These statements do not apply to pediatric practice.
International Relevance
There is no evidence to suggest that the international approach to identification
and management of PAD in people with CKD should differ.
Implications for Clinical Practice and Public Policy
People with CKD are more at risk of PAD and need regular assessment and monitoring.
The role of ankle-brachial index versus other diagnostic techniques may have implications
for future practice. Prospective data on non-surgical therapies and data regarding
percutaneous versus surgical revascularization are required to inform policy and recommended
procedure.
Areas of Controversy, Confusion, or Non-consensus
There remains ongoing debate as to the timing and use of specific diagnostic criteria
for identification of PAD in people with CKD and no studies have examined the utility
of conventional methods in early detection of PAD in CKD. The use of specific medical
and surgical management strategies in people with CKD and PAD has not been evaluated.
There is no reason to believe that treatment strategies should differ, though risks
of diagnostic testing (such as angiography) remain real.
4.4 MEDICATION MANAGEMENT AND PATIENT SAFETY IN CKD
Many drugs and investigative pharmaceuticals are renally excreted and their dosage
may need to be reduced in patients with CKD in order to avoid toxicity. As many people
with CKD are elderly, there should also be consideration of chronological age.
It is beyond the scope of this guideline to list all the agents that may need dose
adjustment in CKD or which should be avoided due to potential nephrotoxicity. This
information is widely available in documents which may exist at local, regional, or
national bodies (e.g., British National Formulary: www.bnf.org), and other textbooks
of pharmacology.
593, 594
However key classes of commonly prescribed drugs in people with CKD will be mentioned
together with suggestions for dose adjustment. Much of this guidance is based upon
our understanding of pharmacology and pharmacokinetics rather than randomized control
trial evidence.
The statements presented here are intended to inform clinicians caring for those with
CKD or at risk for CKD with respect to common clinical situations or exposures which
may put people at risk for AKI or progression of CKD.
4.4.1: We recommend that prescribers should take GFR into account when drug dosing.
(1A)
4.4.2: Where precision is required for dosing (due to narrow therapeutic or toxic
range) and/or estimates may be unreliable (e.g., due to low muscle mass), we recommend
methods based upon cystatin C or direct measurement of GFR. (1C)
4.4.3: We recommend temporary discontinuation of potentially nephrotoxic and renally
excreted drugs in people with a GFR <60 ml/min/1.73 m2 (GFR categories G3a-G5) who
have serious intercurrent illness that increases the risk of AKI. These agents include,
but are not limited to: RAAS blockers (including ACE-Is, ARBs, aldosterone inhibitors,
direct renin inhibitors), diuretics, NSAIDs, metformin, lithium, and digoxin. (1C)
4.4.4: We recommend that adults with CKD seek medical or pharmacist advice before
using over-the-counter medicines or nutritional protein supplements. (1B)
4.4.5: We recommend not using herbal remedies in people with CKD. (1B)
4.4.6: We recommend that metformin be continued in people with GFR≥45 ml/min/1.73
m2 (GFR categories G1-G3a); its use should be reviewed in those with GFR 30–44 ml/min/1.73
m2 (GFR category G3b); and it should be discontinued in people with GFR <30 ml/min/1.73
m2 (GFR categories G4-G5). (1C)
4.4.7: We recommend that all people taking potentially nephrotoxic agents such as
lithium and calcineurin inhibitors should have their GFR, electrolytes and drug levels
regularly monitored. (1A)
4.4.8: People with CKD should not be denied therapies for other conditions such as
cancer but there should be appropriate dose adjustment of cytotoxic drugs according
to knowledge of GFR. (Not Graded)
RATIONALE
These statements are worded in this way in order to inform appropriate prescribing
and management of medical conditions in people with CKD. Recommendation 4.4.1 is specifically
worded to ensure that clinicians remember that in those circumstances where accurate
GFR is required (narrow therapeutic or toxic window), a direct measurement, not an
estimate of GFR should be undertaken.
Evidence Base
Potential problems associated with use of medication in people with CKD include:
Reduced ability to excrete drugs and/or their metabolites
Increased sensitivity to medications (e.g., those bound to albumin in hypoalbuminemic
states such as nephrotic syndrome)
Diminished tolerance of side effects, particularly in the elderly
Loss of efficacy13
Wherever possible, people with CKD should receive the same treatment as those with
normal renal function. However, the dosages may need adjustment according to GFR.
There are medicines whose toxicity is worsened in acute illness particularly in a
setting of dehydration such as diarrhea and vomiting. General advice about appropriate
dosing and when to restart these agents should be given to people taking these drugs
during intercurrent illness, together with a recommendation for consultation with
a health-care professional as soon as possible.
Use of herbal and over-the-counter medicines is very common worldwide and some (such
as those containing aristolochic acid)
595, 596
are known to be nephrotoxic. There is no good quality safety or efficacy data for
many of these compounds.
Table 32 details specific advice for various classes of drugs where CKD may be an
issue. These include RAAS blocking agents,262 beta-blockers, analgesics, antibiotics,
lithium,
597
hypoglycemic agents,
598, 599, 600
lipid-lowering agents,
449, 601, 602, 603, 604
certain chemotherapeutic agents, and anticoagulants.
605
International Relevance
This guidance is based upon knowledge of pharmacology that has universal relevance.
The main international implication is centered on costs of some newer therapies compared
to the older ones. Somewhat paradoxically, as the weight of evidence resides mainly
with agents that have been available for longer, they have the advantage of being
less costly, and have the side effects that are well documented; thus many of these
older agents are preferred.
Pediatric Considerations
No specific recommendation for or against the use of a BSA adjusted eGFR in relation
to drug dosing can be made in children with CKD as the majority of drug studies have
been performed in adult males and extrapolated to children without consideration for
BSA and/or pediatric renal function or clearance.
The exception to this rule would be that for any drug where pediatric pharmacokinetic
or pharmacodynamic studies exist, the method used in calculating the effect of renal
function should be used when estimating the need for dose adjustments or modifications
for the individual patient.
Numerous resources for dose adjustments for pediatric CKD and ESRD patients exist,
including those found in numerous textbooks
606, 607
and online resources.
608
An additional value of dedicated pediatric renal multidisciplinary teams, in terms
of caring for children with CKD and ESRD, is their ability to provide anticipatory
guidance for the patients and families. All such teams should at the least provide
written or online information to their patients and families directing them to seek
advice in situations where they may be prescribed medications from other providers
or may be seeking over-the-counter drugs or supplements. The presence of trained pediatric
pharmacists as part of such teams can be invaluable in achieving such support and
education for both families as well as the CKD health-care providers and community
resources such as local pharmacists.
LIMITATIONS
People with CKD are often excluded or not identified from trials of medications for
non-renal disease. Recommendations are partly based upon knowledge of pharmacology
rather than controlled trials in carefully defined populations.
RESEARCH RECOMMENDATIONS
National and international research groups, and those with CKD-focus organizations
(International Society of Nephrology, International Federation of Kidney Foundations,
and other national bodies) should ensure adequate representation of people with CKD
in clinical trials, leading to an improved understanding of pharmacodynamics of those
with CKD.
4.5: IMAGING STUDIES
4.5.1: Balance the risk of acute impairment in kidney function due to contrast agent
use against the diagnostic value and therapeutic implications of the investigation.
(Not Graded)
RATIONALE
Use of iodinated radiocontrast media has been associated with AKI with reported rates
of 0-11% depending on the population under study, the type of agent that is used,
and the definition of nephrotoxicity.
609
The following recommendations are concordant with those from the American College
of Radiology (ACR),
610
the European Society of Urogenital Radiology (ESUR)
611
and the KDIGO Clinical Practice Guideine for Acute Kidney Injury.7
Radiocontrast
4.5.2: We recommend that all people with GFR <60 ml/min/1.73m2 (GFR categories G3a-G5)
undergoing elective investigation involving the intravascular administration of iodinated
radiocontrast media should be managed according to the KDIGO Clinical Practice Guideline
for AKI including:
Avoidance of high osmolar agents (1B);
Use of lowest possible radiocontrast dose (Not Graded);
Withdrawal of potentially nephrotoxic agents before and after the procedure (1C);
Adequate hydration with saline before, during, and after the procedure (1A);
Measurement of GFR 48–96 hours after the procedure (1C).
RATIONALE
These statements have been worded in order to inform safe radiological investigation
of people with CKD and to avoid potential nephrotoxcity that can be associated with
radiological imaging.
Evidence Base
Radiocontrast media associated AKI is a largely preventable cause of morbidity and
mortality. There is no internationally agreed definition but most studies use an increase
in SCr of >0.5 mg/dl (44 μmol/l) and/or a 25% increase from baseline SCr within 3
days of the procedure.7,
609, 610, 611
Epidemiological studies and case series have identified the following risk factors
for AKI:
GFR <60 ml/min/1.73 m2 (particularly if <30 ml/min/1.73 m2)
Diabetes
Concurrent dehydration
CHF
>70 years of age
Concurrent use of known nephrotoxic agents such as NSAIDs
Use of high osmolality agents (especially in those with GFR <60 ml/min/1.73 m2)
Use of large doses of radiocontrast media
i) Intra-arterial injection
Gout (hyperuricemia)
Numerous studies of preventative strategies have been performed with the following
conclusions:
High osmolar agents pose a greater risk of AKI in people with CKD.
609
Iso-osmolar agents compared to low-osmolar agents are associated with lower rates
of AKI in some but not all studies. Wherever possible iso-osmolar agents should be
used in people with CKD at high risk for AKI (although these tend to be more expensive).
612
Although risk for AKI increases with GFR <60 ml/min/1.73 m2, the rates are particularly
high (7.8% in one study) when GFR is <30 ml/min/1.73 m2. Implementing preventative
strategies for all with a GFR <60 ml/min/1.73 m2 may not be practical but a graded
risk assessment taking into account all factors may be more realistic. Some guidelines
such as American College Radiology
610
and ESUR
611
provide a checklist as a means of identifying patients at risk of AKI before investigation.
Extracellular volume expansion is widely recommended although there are few good quality
trials on which to base an ideal protocol.
609
0.9% saline given by continuous infusion appears superior to 0.45% saline or bolus
injection and there is no demonstration of consistent superiority of sodium bicarbonate
over saline. Current guidance suggests either infusion of 1 ml/kg body weight/hour
for 3-12 hours before and after the procedure or 100 ml/hr, beginning 6 to 12 hours
before and continuing 4 to 12 hours after intravascular iodinated contrast medium
administration.
609, 610
Use of N-acetylcysteine or ascorbic acid as preventative measures has not been shown
to be a consistent benefit.
International Relevance
This guidance has universal relevance although there are cost implications as the
iso-osmolar contrast media are more expensive.
LIMITATIONS
Some of the guidance is based upon limited evidence and there is a need for more research
into simple preventative measure such as pre-investigation rehydration (see below).
There has not previously been a universal definition for AKI following administration
of contrast media. However, recommendations from the KDIGO AKI Guideline suggest that
the same general AKI definition and staging be used for changes in kidney function,
irrespective of etiology.
RESEARCH RECOMMENDATIONS
Prospective studies using direct measures of GFR before and after administration of
radiological contrast media are required to help define the incidence of AKI. Such
studies would also be able to validate creatinine or other estimates of GFR in people
undergoing radiological investigation.
Prospective controlled trials of rehydration using different fluids (saline, bicarbonate,
Hartmann's) and validated estimates of GFR are urgently required.
Definitive studies of N-acetylcysteine and other antioxidants would help determine
their usefulness or otherwise.
Gadolinium-containing contrast media
Gadolinium is a rare earth element that is naturally highly toxic. When bound to proprietary
chelating agents, it is essentially biologically inert in people with normal renal
function and provides excellent contrast during MRI. These chelates are excreted unchanged
by the kidneys by glomerular filtration and have much lower direct nephrotoxicity
than conventional radioiodine contrast media.
613
However over 200 cases of a scleroderma-like condition now termed nephrogenic systemic
fibrosis (NSF) following gadolinium use in patients with CKD have been reported.
613, 614
4.5.3: We recommend not using gadolinium-containing contrast media in people with
GFR <15 ml/min/1.73 m2 (GFR category G5) unless there is no alternative appropriate
test. (1B)
4.5.4: We suggest that people with a GFR <30 ml/min/1.73 m2 (GFR categories G4-G5)
who require gadolinium-containing contrast media are preferentially offered a macrocyclic
chelate preparation. (2B)
RATIONALE
These statements have been worded in order to enable safe administration of gadolinium
to people with CKD. As with all tests requested in the CKD population, clinicians
should be aware of the risk-benefit ratio of the use of gadolinium in people with
GFR 15-29 ml/min/1.73 m2.
Evidence Base
NSF is an untreatable and sometimes fatal condition that complicates the use of gadolinium-containing
contrast media in people with CKD. Prevention is therefore the best approach with
avoidance of gadolinium exposure unless clinically indicated and to use the lowest
risk agent at the lowest dose.
613, 615
Although the risk of AKI with iodinated contrast media is much greater than that for
NSF with gadolinium, the former is treatable with dialysis whereas the latter is not.
Thus, consideration of more conventional imaging techniques should be undertaken in
all with a GFR <30 ml/min/1.73 m2.
610, 613, 614, 615
A recent meta-analysis and review have highlighted those patients most at risk and
quantified an OR for NSF of between 20-50 for those with a GFR <15 ml/min/1.73 m2.
614
While there is general agreement that patients with a GFR >15 ml/min/1.73 m2 are at
increased risk, it is not possible to derive a precise estimate. No patients with
a GFR >30 ml/min/1.73 m2 have developed NSF without concomitant liver failure.
614
There is some suggestion that the type of gadolinium preparation plays a role:
614, 615
Linear chelated preparations such as gadodiamide may be more likely to cause NSF and
should be avoided when GFR is <30 ml/min/1.73 m2. Gadoteridol, gadobutrol, or gadoterate
should be considered if MRI with contrast is clinically essential.
610, 613, 615
Moreover, as gadolinium is freely dialysed,
616
most guidelines recommend dialysis in patients with a GFR <15 ml/min/1.73 m2 or for
those already on dialysis immediately after (and perhaps repeated 24 hours later)
completion of the procedure.
610, 615
The role of dialysis in people with GFR above 15 ml/min/1.73 m2 is uncertain.
International Relevance
This guidance has universal relevance although there are cost implications as the
non-linear chelated preparations are more expensive.
Pediatric Considerations
Regarding both Recommendations 4.5.3 and 4.5.4, a number of considerations specific
to the use of gadolinium preparations in young children and neonates must also be
considered in addition to the general admonishments against their use in situations
of GFR <30 ml/min/1.73 m2. In particular, the FDA currently does not license any gadolinium-based
contrast agent (GBCA) product for use in children less than 2 years of age; and likewise
the European Medicines Agency cautions against the use of any GBCA in a child less
than 1 year of age.
In addition, the ability to accurately estimate the relative value of a neonate or
young infant's GFR leads to great difficulty in terms of assigning risk of GBCA exposure
if one bases this on renal clearance. In a recent paper by Meng and colleagues,
617
they surveyed a worldwide group of both cardiologists and radiologists with respect
to their use of gadolinium in their pediatric MRI practice. While 93% of the respondents
did evaluate renal function in some or all of their patients, only a slight majority
(54%) used an estimating formula, most often the Schwartz equation, with the remainder
relying on SCr alone (31%) or urine output (6%). Perhaps of most concern was that
renal function was assessed in only 33% and 31%, respectively, of patients classified
as ‘all neonates' or ‘all patients <1 week of age.' Equally concerning was the fact
that 13% of the respondents would give gadolinium to some of these children in the
face of a GFR <30 ml/min/1.73 m2.
In recognition of our inability to accurately measure GFR in the neonate and, by extension,
the clearance of compounds such as gadolinium, all nephrologists and radiologists
must exercise caution in terms of use of GBCA in this potentially high-risk population,
and all other imaging modalities should be considered prior to choosing one requiring
gadolinium exposure.
LIMITATIONS
The evidence is largely based upon case series rather than prospective studies or
RCTs.
RESEARCH RECOMMENDATIONS
As for iodinated contrast media, a prospective study of people with CKD undergoing
nuclear MRI with gadolinium contrast would help define change in GFR and validate
estimators. Because NSF is such a serious condition, an RCT of dialysis in people
with GFR <30 ml/min/1.73 m2 would help to determine risk-benefit in these patients,
though recruitment may be difficult due to potential ethical concerns.
Bowel preparation
The increasing use of colonoscopy as a screening tool for bowel cancer has resulted
in many people undergoing bowel preparation with oral sodium phosphate-containing
preparations. Case reports of acute and late irreversible renal failure with biopsy-proven
phosphate deposition have led to a new disease entity termed acute phosphate nephropathy.
618, 619
Less than 40 definite cases have been reported and these have been extensively reviewed.
4.5.5: We recommend not to use oral phosphate-containing bowel preparations in people
with a GFR <60 ml/min/1.73 m2 (GFR categories G3a-G5) or in those known to be at risk
of phosphate nephropathy. (1A)
RATIONALE
This statement has been worded in order to enable safe bowel preparation in people
with CKD who need to undergo investigation of bowel disease.
Evidence Base
Electrolyte disturbances that are sometimes severe and include hyperphosphatemia,
hypocalcemia, hypo- and hypernatremia, and hypokalemia have been reported in normal
volunteers undergoing oral phosphate bowel preparation.
620
Renal injury has been reported in a small number of people although the condition
is likely to be under-recorded. A recent study from Iceland estimated the incidence
to be around 1 per thousand doses
621
but others would suggest incidence rates of between 1% and 4%.
618, 619
Two broad patterns of renal injury have been described. An early symptomatic response
associated with severe hyperphosphatemia and hypocalcemia and a later (days to months)
irreversible kidney injury associated with a specific tubulointerstitial calcium phosphate
deposition.
618, 619
The following people are said to be at particular risk although the link to kidney
injury is associative in many cases and firm evidence is lacking:
GFR <60 ml/min/1.73 m2
>60 years of age
Female
Hypertension
Diabetes
CHF
Dehydration
Active colitis
Concurrent use of RAAS blocking agents, diuretics, lithium, NSAIDs
Large and/or repeat dosing of oral phosphate preparations
Hypoparathyroidism
Although the FDA has banned oral phosphate solutions, there is no qualitative difference
with tablet preparations. Both should be avoided in people at risk and this is the
current recommendation of the American Society for Gastrointestinal Endoscopy.
622
There is some debate as to whether the cause of the kidney injury is entirely due
to dehydration rather than phosphate use per se.
623
However, it is hard to understand why calcium phosphate deposits occur and cause late
renal impairment if dehydration was the only factor.
There is a single RCT of phosphate versus non-phosphate preparations but assessment
of renal function was limited to change in SCr and GFR at baseline is not reported.
624
Notwithstanding these limitations, there were greater changes in serum potassium,
calcium, and phosphate in those given sodium phosphate-containing preparations.
As there are non-phosphate-containing bowel preparations available, these should be
used in all the above groups (and arguably in all people given the biochemical abnormalities
observed in normal volunteers). As for radiocontrast media, rehydration with saline
may be required in the frail and ill irrespective of the bowel preparation that is
used.
International Relevance
This guidance has universal relevance but non-phosphate-containing bowel preparations
are more expensive, so the use of these agents may differ around the world.
LIMITATIONS
The data are based upon case series with a limited number of affected individuals
and retrospective population studies.
RESEARCH RECOMMENDATIONS
Prospective study in people with normal renal function and those with different severities
of CKD are urgently required in order to define the acute biochemical and metabolic
effects of phosphate-containing bowel preparations. There is also a need for a study
of all people undergoing bowel preparation with whatever preparation in order to explore
the effects on GFR on the incidence of this complication. More definitive exploration
of rehydration therapy (type and volume) in people with CKD undergoing bowel preparation
is urgently needed.
4.6 CKD AND RISKS FOR INFECTIONS, AKI, HOSPITALIZATIONS, AND MORTALITY
The following section gives guidance for the care of people with CKD given that they
are at increased risk for a number of events such as infections, AKI, CVD, hospitalizations,
and mortality. Appreciating that increased risk and implementing some of the recommendations
below may result in improved outcomes for people. It will be important to develop
policies and robust research agendas to address areas which do not have a substantial
evidentiary base.
CKD and risk of infections
CKD is associated with significant major infectious complications, which occur at
rates 3 to 4 times the general population. Infection is an important cause of morbidity
and mortality among patients with kidney failure and is the second leading cause of
death following CVD. CKD may be a risk-multiplier for acute infectious disease-associated
mortality, as it is for CVD. Despite a less effective response to vaccination, there
are data suggesting benefit from immunization among people with CKD as in the general
population.
4.6.1: We recommend that all adults with CKD are offered annual vaccination with influenza
vaccine, unless contraindicated. (1B)
4.6.2: We recommend that all adults with eGFR <30 ml/min/1.73 m2 (GFR categories G4-G5)
and those at high risk of pneumococcal infection (e.g., nephrotic syndrome, diabetes,
or those receiving immunosuppression) receive vaccination with polyvalent pneumococcal
vaccine unless contraindicated. (1B)
4.6.3: We recommend that all adults with CKD who have received pneumococcal vaccination
are offered revaccination within 5 years. (1B)
4.6.4: We recommend that all adults who are at high risk of progression of CKD and
have GFR <30 ml/min/1.73 m2 (GFR categories G4-G5) be immunized against hepatitis
B and the response confirmed by appropriate serological testing. (1B)
4.6.5: Consideration of live vaccine should include an appreciation of the patient's
immune status and should be in line with recommendations from official or governmental
bodies. (Not Graded)
4.6.6: Pediatric immunization schedules should be followed according to official international
and regional recommedations for children with CKD. (Not Graded)
RATIONALE
CKD is associated with alterations in primary host defense mechanisms and increases
the risk of bacterial infections (Table 33). Epidemiological study suggests that the
3 most commonly seen infectious complications in the CKD population are: urinary tract
infection, pneumonia, and sepsis. In the general population, there is strong evidence
that preventive measures are effective in adults and there are data suggesting benefit
from immunization in people with CKD.
Evidence Base
Previous investigations have firmly established ESRD as a strong risk factor for infectious
complications.
625, 626, 627, 628
However, few epidemiologic reports have addressed the risk of infections in people
with CKD not treated with dialysis.
629, 630
Data from the US Renal Data System (USRDS) suggest that higher rates of hospital admission
because of septicemia are noted in people with CKD compared with those without CKD.
631
Comprehensive studies of the absolute rates, risk factors, clinical course, and outcomes
of different types of clinically relevant infections across the spectrum of CKD are
not available.
Among people with CKD, very few studies have examined the incidence or prevalence
of infections and no published data describe GFR category–specific infection rates.
Among Medicare beneficiaries aged more than 66 yr, people with diagnosed CKD seem
to have substantially higher rates of hospitalization for diagnoses of pneumonia and
sepsis compared with people without diagnosed CKD.
628
In addition to an increased incidence of being hospitalized with infections, people
with CKD have longer lengths of hospital stay during infection-related admissions
compared with people without CKD.
632
Among USRDS data admission rates for all causes, CVD and infection, are 38–46% higher
for those with CKD than for those without.
633
Rates for pneumonia in CKD were nearly three times higher than those for non-CKD.
Hospitalizations for bacteremia/septicemia, were nearly four times higher for people
with CKD compared to those without. Hospitalization rates for urinary tract infections
were three-fold higher for people with CKD.
In people with ESRD, the function of polymorphonuclear white blood cells, lymphocytes,
and monocytes is altered, resulting in an impaired host response to infection.
636, 637, 638
However, these issues have not been adequately investigated in people with CKD.
There is a growing body of evidence in the general population for reduction in infections,
hospitalizations, and mortality as a result of immunizations. Current data suggest
that vaccination is an underused prevention strategy in the CKD and ESRD populations.
639
Potential barriers to immunization that are specific to the CKD populations have not
been systematically examined. Although lower vaccine responsiveness has been widely
recognized in the ESRD population, to what extent moderate to advanced CKD modifies
vaccine responsiveness remains unclear.
640, 641, 642
Studies that have examined vaccination in the setting of CKD or ESRD have been limited
by small study size, variable follow-up, and ascertainment of surrogates for vaccine
effectiveness such as antibody response and rate of antibody decline after vaccination
as opposed to vaccine efficacy for preventing infection.
643
Influenza A and B vaccine.
Despite potentially impaired antibody responses, a 2-year analysis of US Medicare
claims data found that people vaccinated against influenza A and B on dialysis had
a substantially lower chance of any-cause hospital admission and any-cause death than
those patients not vaccinated on dialysis.
644
This finding might indicate clinical effectiveness of vaccinating this population
but its observational design might also reflect differences in the underlying clinical
status among people vaccinated and not vaccinated. No unique adverse events related
to influenza vaccine have been identified in people on dialysis.
Pneumococcal vaccine.
People with kidney disease vaccinated with the pneumococcal vaccine seem to develop
different serotype-specific titers, develop lower levels of antibody titers, and have
a more rapid loss of antibody titers as compared with healthy control subjects.
639, 645, 646
Practitioners should be aware of the impact of specific vaccines on responsiveness
and duration of responsiveness. Revaccination practices will be dictated by that knowledge.
Hepatitis B vaccine.
Widespread hepatitis B virus (HBV) vaccination at the onset of dialysis has led to
a marked reduction of HBV infections in people with ESRD, although improved screening
of blood products and dissemination of recommendations for reducing the spread of
HBV infections in dialysis units have also likely contributed.
647
Among people with moderate to advanced CKD, hepatitis B vaccination responsiveness
has been shown to range from approximately 60 to 80% depending on the dosage, number
of administered vaccines, and study population. Although findings have been inconsistent
as to whether the level of GFR affects vaccine responsiveness in people with CKD
640, 642
those with higher GFR are more likely to respond with seroconversion, independent
of other factors.
640
S. aureus vaccine.
StaphVAX has not been shown to be efficacious in reducing the risk for Staphylococcus
aureus bacteremia in people on hemodialysis.
648, 649
No data are published in people with CKD.
Live vaccines.
Due to the fact that people with CKD are often immunocompromised, live vaccines should
only be used with caution on an individual basis.
In summary, although some vaccines (like influenza) in usual doses provide protection,
other vaccines (HBV and pneumococcal) require more frequent dosing or larger doses
to achieve and maintain protective antibody titers. Frequency and type of vaccination
will vary according to local circumstances and prevalence of disease.
International Relevance
The availability of different vaccinations may vary worldwide, as does the prevalence
of specific bacterial, viral, and other infections. It is reasonable to offer individuals
appropriate immunization according to local practices.
Implications for Clinical Practice and Public Policy
Vaccines for influenza, hepatitis B, and pneumococcus are currently recommended for
people with CKD by the local, regional, or national advisory committees on immunization
practices from most countries.
Current recommendations are to:
Provide influenza vaccination annually to people with CKD.
Provide pneumococcal vaccine with a single booster dose 5 years after the initial
dose.
Provide HBV vaccine to people with CKD who are likely to require RRT. Although the
recommendation is to give the HBV vaccine during more severe CKD (GFR <15 ml/min/1.73
m2), it may be preferable to give this earlier to maximize the chances of achieving
immunity; there are data to support this practice.
640
This would also ensure that all patients are immunized against HBV before receiving
a transplant. As protective antibody levels may fall, this should be checked (possibly
annually) with booster doses given if appropriate.
Areas of Controversy, Confusion, or Non-consensus
Much remains to be understood concerning impaired host response to infection in patients
with CKD.
Studies should be undertaken to determine the absolute rates, risk factors, and clinical
course of different types of clinically relevant infections across the spectrum of
CKD, by GFR and albuminuria category, and by cause.
The outcomes across the range of acute infections in CKD population need to be ascertained.
Studies should be undertaken to assess the rate of decline of antibody titers post-vaccination
and the efficacy of immunization in people with CKD.
Pediatric Considerations
Current immunization schedules for children are regularly updated by both the US Centers
for Disease Control and Prevention
650
and American Academy of Pediatrics.
651
Current and comprehensive immunization recommendations for children with CKD have
been published by Neu in 2012.
652
The paper addresses key issues regarding the use of vaccines in CKD pediatric populations
who are receiving concomitant immunosuppression and in those awaiting transplantation.
The need for, and interpretation of, protective antibody levels for those vaccines
where this is indicated is described.
An oversimplified summary of the recommendations would be to provide all recommended
childhood vaccines to every child with CKD with the exception of any live viral vaccine
in a child receiving immunosuppressive medications. Likewise children on dialysis
should not receive the live attenuated influenza vaccine although the inactivated
version can and should be given to all children with CKD on an annual basis otherwise.
Pneumococcal vaccination is particularly important in children with nephrotic syndrome
and those with CKD, and current vaccination schedules and products should be carefully
reviewed to ensure proper serotype coverage is being provided. Hepatitis B status
and vaccination are of extreme importance in all children who may go onto dialysis
– and specific recommendations for ongoing monitoring and interpretation of antibody
levels should be carefully reviewed.
CKD and risk of AKI.
Due to the epidemiological association between CKD and AKI and the number of observational
studies reporting an association between pre-existing CKD and AKI, CKD is considered
the most consistent pre-existing condition associated with a high risk of AKI. However
the potential linkage between patients with AKI, CKD, and ESRD has been inadequately
studied to date and remains ill defined. This section describes AKI as a complication
which needs to be managed in those with CKD. Given its association with progression,
it is also described in that section.
4.6.7: We recommend that all people with CKD are considered to be at increased risk
of AKI. (1A)
4.6.7.1: In people with CKD, the recommendations detailed in the KDIGO AKI Guideline
should be followed for management of those at risk of AKI during intercurrent illness,
or when undergoing investigation and procedures that are likely to increase the risk
of AKI. (Not Graded)
RATIONALE
Observational data suggest a strong association between pre-existing CKD and AKI.
The appreciation that CKD patients may be more susceptible to AKI is the purpose of
the above set of statements. However, methodological issues such as how CKD and AKI
are defined in clinical studies and the statistical adjustments for non-uniformity
of comorbidities among various studies may affect the validity of observed associations.
Evidence Base
CKD is designated as a risk factor for AKI because of the epidemiological association
between the two.263,264 A number of studies in a variety of settings report an association
between pre-existing CKD and AKI. 265-271 CKD is a potent predictor of acute decline
in kidney function following exposure to radiocontrast,272 major surgery,273 and other
medical conditions.274
Hsu et al.14 compared the pre-hospitalization MDRD GFR of 1764 adult members of the
Kaiser Permanente Northern California health-care system who developed dialysis-requiring
AKI during hospitalization with 600,820 individuals who did not. Compared with a reference
baseline GFR of ≥60 ml/min/1.73 m2, a baseline GFR of 45–59 ml/min/1.73 m2 was associated
with an adjusted OR of in-hospital AKI of 1.66 (95% CI 1.40–1.97). For GFR values
of 15–29 ml/min/1.73 m2, the adjusted OR for in-hospital AKI was 20.42 (95% CI 17.40–23.96).
The presence of diabetes, hypertension, and proteinuria increased the likelihood of
developing in-hospital AKI, with adjusted ORs of 1.99 (95% CI 1.78–2.23), 1.55 (95%
CI 1.37–1.76) and 2.84 (95% CI 2.52–3.19), respectively. The authors concluded that
CKD is the main risk factor for AKI during hospitalization. A contrasting approach
by Singh et al. defined AKI as dialysis-requiring acute renal failure.275 Because
the clinical decision to dialyze a patient is frequently influenced by a higher overall
SCr, presence of hemodialysis access, or consideration of inevitable progression to
ESRD, this definition of AKI could bias toward capturing more AKI cases in CKD patients.
Moreover, in patients with advanced CKD, the progression of CKD to ESRD may sometimes
be difficult to separate from acute-on-chronic renal failure. A cohort study by Lafrance
et al. followed a referred CKD population in British Columbia for a median of 19.4
months after achieving a GFR of ≤30 ml/min/1.73 m2. Forty-five percent had at least
one episode of AKI.276 In another cohort study of 920,985 adults in Alberta, Canada
with at least one outpatient measurement of SCr and proteinuria and not requiring
chronic dialysis, risk of admission with AKI increased with heavier proteinuria and
reduced GFR.16
International Relevance
These guidelines about AKI have relevance around the world. While the causes of AKI
may differ by region, country, socio-economic status, and age, the consequences remain
the similar. Where there are no facilitities to support AKI or CKD, people will die.
Areas of Controversy, Confusion, or Non-consensus
Interpretation of published data examining the influence of pre-existing CKD on the
increased likelihood of AKI is potentially confounded by a number of issues. These
include the comorbidities associated with CKD, influenced by repeated exposure to
various nephrotoxic insults or in-hospital errors,57,277 or primarily due to the altered
physiology in CKD. There are also methodological issues such as how CKD and AKI are
defined in clinical studies and the varying statistical adjustments for comorbidities
which may affect the validity of observed associations.
A further important issue to clarify is whether pre-existing CKD influences the outcome
of AKI. Currently, there is no single biomarker that can differentiate ‘acute' from
‘chronic' kidney disease and help to address this issue. Several large observational
and database studies report, surprisingly, lower in-hospital mortality in patients
with AKI superimposed on CKD compared with controls.278-283 Data from PICARD reveal
lower in-patient mortality and median length of stay in ICU subjects with acute-on-chronic
renal injury compared with non-CKD subjects with AKI, though the post-discharge dialysis
rates were higher in subjects with pre-existing CKD.284
Pediatric Considerations
The relative paucity of pediatric specific guidelines (due to lack of high-quality
studies) in the KDIGO AKI guideline would suggest that the use of pediatric data and
review papers as well as relevant pediatric nephrology texts would be of benefit to
the practitioner interested in reviewing this topic in greater detail and applying
pediatric data to their practice.
653, 654, 655, 656, 657, 658
RESEARCH RECOMMENDATIONS
Prospectively designed clinical studies with a clear and uniform definition of CKD
and AKI and adjusted for comorbidities are needed to determine:
the frequency of AKI events in a CKD population
the outcome of AKI in patients with CKD condition
the importance of proteinuria in addition to low GFR in the risk of AKI
CKD and risk of hospitalization and mortality
Regardless of the method used to estimate GFR, hospitalization and mortality rates
are higher in people with CKD. Exact rates vary with comorbidity and severity of CKD,
and are not well-defined. Selection of interventions that could reduce hospitalizations,
morbidity, mortality, and costs in people with CKD is not well-studied.
4.6.8: CKD disease management programs should be developed in order to optimize the
community management of people with CKD and reduce the risk of hospital admission.
(Not Graded)
4.6.9: Interventions to reduce hospitalization and mortality for people with CKD should
pay close attention to the management of associated comorbid conditions and cardiovascular
disease in particular. (Not Graded)
RATIONALE
There are observational and database studies reporting an association between pre-existing
CKD and hospitalizations and mortality. A better understanding of the rates, causes,
and risk factors for hospitalization among people with CKD would allow estimates of
the economic burden of CKD and identification of those at risk for increased resource
utilization. People with CKD are an ideal target for interventions aimed at reduction
of morbidity, hospitalization, mortality, and costs. These statements suggest that
a coordinated approach to the identification and management would result in better
outcomes and are intended as ‘best practices' statements, recognizing the difficulty
in developing an evidence base while addressing issues related to resource allocations
(Table 34).
Evidence Base
Mortality rates remain high (16-22%) with the use of dialysis, with more than half
of all deaths related to CVD. Less is known about mortality and CVD rates, and resource
use among persons with a reduced GFR who are not yet receiving maintenance dialysis.
Few studies have investigated the association between CKD and the risk of hospitalization.
Data from the USRDS reveal that hospitalization rates vary with comorbidity and interact
with degrees of CKD. Adjusted rates are 38% higher in people with CKD and 19% greater
in people with CKD and GFR under 60 ml/min/ 1.73 m2 than in those with GFR >60 ml/min/1.73
m2, who in turn are 20% higher than in people without CKD-illustrating the graded
impact of advancing kidney disease. Not surprisingly, rates of cardiovascular hospitalization
are greater for people with CKD, particularly those with increasing severity of CKD.
659
In both the CKD and non-CKD populations, adjusted rates of hospitalization increase
with greater comorbidity. In 2008, for example, the rate for people with CKD with
both diabetes and CHF was 726 per 1000 person-years at risk — 85% greater than the
rate of 393 among people with neither diagnosis. Rates of admission for CVD increase
even more in the higher categories of GFR. Among Medicare subjects, the rate of 141
admissions per 1000 person-years for those with GFR <60 ml/min/1.73 m2 is 26% higher
than the rate of 112 reported for those with CKD and GFR ≥60 ml/min/1.73 m2. The admission
rates of 101 and 90 reported for MarketScan and Ingenix i3 subjects with GFR <60 ml/min/1.73
m2 are 48 and 16% greater, respectively, than those occurring in people with higher
GFRs.
633
Adjusted rates of mortality in USRDS 2008 increased with age, and were highest in
people with advanced categories of GFR: 31–72% higher, for example, in people with
GFR <60 ml/min/1.73 m2 compared to those with no CKD. By gender, rates in men with
CKD were 91.8 per 1000 person-years at risk compared to 85.6 in women. Rates for people
with CKD overall were similar in whites and African Americans, but in people with
GFR <60 ml/min/1.73 m2, rates for African Americans were 18% higher than those for
whites, at 95.0 and 80.5 per 1000 person-years, respectively.
633
Khan et al.
660
confirmed that hospital utilization among people with CKD is high. During a median
follow-up of 11.4 months, 47% of subjects had at least one hospitalization and there
were on average 0.96 hospitalizations, 6.6 hospital days, and 4.0 outpatient nephrology
visits per person-year at risk. Cardiac disease/hypertension was the most common primary
diagnosis of hospitalizations and progression of CKD/acute kidney failure was the
most common secondary cause of hospitalization. The authors had previously shown that
the dialysis population at their institution had 2.2 hospitalizations and 14.8 hospital
days per person-year at risk,
659
which was similar to that among US hemodialysis patients between 1996 and 1998, who
had 1.9 hospitalizations and 14 hospital days per person-year at risk.
661
In the general population, there were 0.31 hospitalizations and 1.9 hospital days
per person in 1998.
662
Go et al.58 reported an independent, graded association between GFR and the risk of
death, cardiovascular events, and hospitalization in 1,120,295 adults within a large,
integrated system of health-care delivery in whom SCr had been measured between 1996
and 2000 and who had not undergone dialysis or kidney transplantation. These risks
were evident at GFR <60 ml/min/1.73 m2 and substantially increased at GFR <45 ml/min/1.73
m2.
The higher hospital utilization among people with CKD compared to the general population,
and the similarity in the comorbid conditions and the causes of hospitalization between
people with CKD and ESRD, confirm the hypothesis that the complications and comorbidity
observed in ESRD are manifest earlier in the disease process. Previous studies have
demonstrated an association between age, gender, race, cardiac disease, peripheral
vascular disease, serum albumin and hematocrit levels, and resource utilization among
people on dialysis.
659, 663, 664
Holland et al.
665
identified baseline predictors of first non-elective hospitalization among a retrospective
cohort of 362 predialysis subjects. Multivariate analysis, adjusted for baseline creatinine
concentration, selected advanced age (RR 1.02; 95% CI 1.01–1.03), angina (RR 1.9;
CI 1.37–2.61), peripheral vascular disease (RR 1.55, CI 1.05–2.27), and Hb concentration
(RR 0.99, CI 0.94–0.98) as independent predictors of hospitalization. These comorbid
conditions progressively worsen with advancing kidney disease and result in a substantial
proportion of people having severe complications by the time they come to RRT.
Interventions to reduce hospitalizations and mortality for people with CKD should
pay close attention to the management of associated comorbid conditions and CVD.
666
Closer attention to the management of heart disease in this population could substantially
improve outcomes.
667
The influence of correction of anemia on hospitalization is controversial. Drüeke
et al.
668
randomly assigned 603 subjects with GFR 15-35 ml/min/1.73 m2 and mild-to-moderate
anemia (Hb level: 11.0-12.5 g/dl [110-125 g/l]) to a target Hb value in the normal
range (13.0-15.0 g/dl [130-150 g/l]) or the subnormal range (10.5-11.5 g/dl [105-115
g/l]). There were no significant differences between the two groups in the incidence
of hospital admission (61% and 59%, respectively) or mean duration of hospitalization
for cardiovascular reasons (33.0 and 28.2 days, respectively). Singh et al.
669
studied 1432 people with CKD, 715 of whom were randomly assigned to receive a dose
of epoetin alfa targeted to achieve a Hb level of 13.5 g/dl (135 g/l) and 717 of whom
were assigned to receive a dose targeted to achieve a level of 11.3 g/dl (113 g/l).
The median study duration was 16 months. The primary end point was a composite of
death, MI, hospitalization for CHF (without RRT), and stroke. They observed an increased
risk of the primary composite end point in the high Hb group as compared with the
low Hb group. Death and hospitalization for CHF accounted for 74.8% of the composite
events. The Trial to Reduce Cardiovascular Events with Aranesp Therapy (TREAT)375
study involved 4038 people with diabetes, CKD, and anemia. Subjects were randomly
assigned to treatment with darbepoetin alfa to achieve a Hb level of approximately
13 g/dl (130 g/l) or to placebo, with rescue darbepoetin alfa when the Hb level was
less than 9.0 g/dl (90 g/l). The primary end points were the composite outcomes of
death or a cardiovascular event (nonfatal MI, CHF, stroke, or hospitalization for
myocardial ischemia) and of death or ESRD. Again there were no significant between-group
differences in the outcomes of interest.
International Relevance
It is of benefit for all jurisdictions to appreciate the increase in resource utilization
by CKD populations. Identification of people at risk for increased resource utilization
and economic burden of CKD should result in strategies to attenuate that risk or address
the resource implications. Selection of interventions that could reduce hospitalizations,
morbidity, mortality, and costs in these populations should be evaluated.
DISCLAIMER
While every effort is made by the publishers, editorial board, and ISN to see that
no inaccurate or misleading data, opinion or statement appears in this Journal, they
wish to make it clear that the data and opinions appearing in the articles and advertisements
herein are the responsibility of the contributor, copyright holder, or advertiser
concerned. Accordingly, the publishers and the ISN, the editorial board and their
respective employers, office and agents accept no liability whatsoever for the consequences
of any such inaccurate or misleading data, opinion or statement. While every effort
is made to ensure that drug doses and other quantities are presented accurately, readers
are advised that new methods and techniques involving drug usage, and described within
this Journal, should only be followed in conjunction with the drug manufacturer's
own published literature.