|Values are valid only on day of printing.|
Click CC to turn on closed captioning.
Published: August 2011Print Record of Viewing
Chronic kidney disease (CKD) is a worldwide public health problem. There is a rising incidence and prevalence of kidney failure in the United States and early diagnosis is key. Early stages of kidney disease can be detected through laboratory testing. Accurate estimation of glomerular filtration rate (GFR) is essential for the diagnosis, staging, and management of CKD. A new immunoassay for determination of cystasin C has made it more practical and clinically useful to estimate GFR. A formula that uses both serum creatinine and cystatin C with age, sex, and race may be better than equations that use only 1 of these serum markers and may provide the best estimation of GFR and early kidney disease. Cystatin C may also be useful for detecting those patients with CKD who are at highest risk of complications and progression.
Presenter: John C. Lieske, MD
Welcome to Mayo Medical Laboratories' Hot Topics. These presentations provide short discussions of current topics and may be helpful to you in your practice.
Our presenter for this program is John C. Lieske, MD, Medical Director of the Mayo Clinic Renal Testing Laboratory in the Department of Laboratory Medicine and Pathology and Consultant in the Division of Nephrology and Hypertension at Mayo Clinic in Rochester, Minnesota. Dr. Lieske will describe renal function and chronic kidney disease. He will also discuss laboratory assessment of kidney function and describe a new immunoassay method to measure cystatin C which is practical and clinically useful to estimate glomerular filtration rate (GFR).
After viewing the Hot Topic, we invite you to participate in Beyond Hot Topic. This question and answer session will be posted online approximately 1 month after the Hot Topic presentation is posted. You can submit a question for Dr. Lieske using the information at the end of the presentation. Thank you Dr. Lieske.
I would like to begin this presentation with a brief case discussion. Imagine you are a clinician evaluating a 66-year-old recently retired African American businessman. He and his wife have just purchased their dream retirement home. In general, he has been in very good health. However, he has a history of high blood pressure for approximately 12 years treated with medications, and has had high blood sugars noted for at least five years, although these have never been high enough that he was told that he has diabetes. He tells you that his grandmother was on dialysis and he worries about that possibility for himself. A serum creatinine level currently is 1.5 mg/dL.
There are several key questions this case highlights: What can we tell him about his kidney function at the current time? Is he at risk for kidney failure down the road? What further lab tests could we order to help us answer these questions? And finally, How should we treat him in order to minimize his risk of kidney failure?
At this point let us take a step back and ask ourselves a little bit about kidney function. It is important to remember what it is that the kidneys do for us. The quick answer is that they filter blood in order to remove various toxins and waste products of daily metabolism. It is helpful to have a general idea about the amount of filtrating that the kidneys do on a daily basis. It turns out that this number is about 100 L/ day in a healthy individual. It is also helpful to compare this number to our daily urine output, which is much less. A typical number here would be 1 to 3 L. It is important to note that the kidneys filter almost 100 times as much as the amount of fluid that we need to get rid of and that large amount of filtrate is very highly processed to produce our final urine. That daily filtrate, also known as glomerular filtration rate or GFR, is the single best indicator of kidney function and kidney reserve. When GFR falls below certain thresholds, health is impacted.
Therefore, clinicians need accurate ways to estimate this key number. The most important reason is to screen patients for the presence of chronic kidney disease, so that appropriate treatment can be initiated. If a person is known to have chronic kidney disease, then clinicians will want to monitor GFR over time to determine if kidney function is remaining stable or not. If GFR falls too low, then plans need to be made for renal replacement therapy, such as dialysis or transplantation. Finally, certain drugs are eliminated by the kidney, and doses need to be reduced when GFR falls below certain thresholds. Typically GFR is reported not in liters per day, but in milliliters per minute. Furthermore, it is now standard to normalize this number for body surface size, or 1.73 m2.
There are 2 ways that a person can be diagnosed with chronic kidney disease. The first criterion requires the presence of kidney damage for more than 3 months, as defined by structural or functional abnormalities, with or without any decrease in GFR. This could include pathological abnormalities confirmed on a kidney biopsy. Alternatively, markers other than an elevated creatinine can indicate kidney damage on lab testing. An example would be an abnormal urinalysis. The second criterion for chronic kidney disease is a little simpler. Anyone with a GFR less than 60 mL per minute per 1.73 m2 meets criterion for chronic kidney disease.
Patients with chronic kidney disease are now classified into 5 stages. Persons in stage 1 or 2 have good GFR but other evidence of kidney damage. Persons in stage 3 CKD have moderately reduced GFR, less than 60 mL per minute per 1.73 m2. In general, persons in stage 3 are at risk for further loss of kidney function, although this risk is quite variable. Therefore, it is important to intervene and modify all possible risk factors to reduce this risk to the greatest extent possible. Persons in stage 4 CKD have a GFR less than 30 mL per minute per 1.73 m2. Studies suggest such individuals are at high risk for complications of chronic kidney disease, such as anemia and bone disease. Furthermore, cardiovascular complications are very high in persons with stage 4 CKD. Therefore, it is important to identify these persons in order to treat their complications of kidney disease and modify all possible cardiovascular risk factors. Finally, persons with stage 5 CKD have a GFR less than 15 mL per minute per 1.73 m2. This is typically in the threshold where persons are symptomatic and need to start on a renal replacement therapy, such as dialysis or kidney transplantation.
The laboratory plays a key role in the diagnosis and staging of chronic kidney disease. In order to detect kidney damage independent of GFR, tests include a urinalysis and quantification of protein in the urine. Abnormalities that would indicate CKD include proteinuria, albuminuria, and formed elements such as red cell casts. Sometimes patients have more subtle abnormalities that might be detected on a blood electrolyte panel or more sophisticated testing. Examples include an inability to acidify the urine, called renal tubular acidosis, or an inability to concentrate the urine, called nephrogenic diabetes insipidus. Finally, radiology studies might detect evidence of kidney scarring, also diagnostic of CKD by criterion 1.
As mentioned above, criterion 2 is entirely based on your GFR number. There are 2 methods to establish what a person’s GFR is. We can directly measure GFR. Strategies here would include 24-hour urine for creatinine clearance, or specialized testing such as an inulin clearance or iothalamate clearance techniques. Although these methods are useful in many settings, they are more involved and costly, and not particularly suitable for mass screening or frequent monitoring of patients. Fortunately, methods have been developed to estimate GFR based upon simple blood tests.
Serum creatinine is a tried and true marker of GFR. Creatinine is a byproduct of muscle turnover. In general, production of creatinine is very constant from day-to-day in any given individual. Exceptions would include situations where muscle is rapidly damaged such as rhabdomyolysis, or individuals that greatly increase or decrease muscle mass over time. Creatinine is a very small molecule that is freely filtered in the kidney. Hence, clearance of creatinine is proportionate to GFR. However, it is well established that a small amount of creatinine is also secreted by kidney tubules. Therefore, creatinine clearance will always overestimate GFR to some extent. Furthermore muscle mass varies greatly between individuals depending on their age, gender, and size.
This slide clearly demonstrates both the utility and problems with using serum creatinine alone to estimate GFR. Serum creatinine numbers are plotted against measured GFR in a large number of individuals with and without kidney disease. In general, it is quite true that serum creatinine increases as GFR falls. However, creatinine does not increase very much until GFR is well below 60 mL per minute per 1.73 m2. Such a person already would have stage 3 chronic kidney disease. The dotted line corresponds to a serum creatinine of 1.2 milligrams per deciliter, the upper limit of the reference range. Such a serum creatinine could correspond to a GFR as low as 25 or as high as 130 mL per minute per 1.73 m2.
Therefore, efforts have been devoted to make serum creatinine a better estimate of GFR. The main factor that varies between individuals and influences serum creatinine independent of GFR is muscle mass, which in turn is the key determinant of creatinine production. Therefore, equations have been developed that estimate what a person’s creatinine production should be. The oldest example of such an equation is the Cockroft-Gault equation. This equation estimates creatinine clearance based upon serum creatinine together with age, gender, and weight. The MDRD estimated GFR, or eGFR, was developed using data from the Modification of Diet in Renal Disease study. This study only included patients with chronic kidney disease, who were placed on various diets in order to determine if a low-protein diet helped preserve kidney function. As part of the study, patient’s had their GFR measured using iothalamate clearance. The investigators then used data in their study to develop an equation that best estimated GFR. One problem with the MDRD equation is that it was developed using only persons with CKD, and therefore does not perform very well in normal individuals. Therefore, a newer equation has been developed using a mixed population of patients with and without CKD. This equation is called the CKD-EPI equation. So far it is used mostly for research, but may make it into clinical practice sometime in the future.
This slide shows the eGFR equation that was developed in the MDRD study, and is most commonly reported by lab information systems. The key demographic variables are age, gender, and race. Shown is the equation one would use for serum creatinines measured by an enzymatic creatinine assay that is traceable to isotope dilution- mass spectrometry standards. There is a slightly different equation that would be applied if a lab was using certain Jaffe creatinine assays. Since the MDRD eGFR equation does not perform well in persons with normal kidney function, most labs do not report an actual number unless it is less than 60 mL per minute per 1.73 m2. Furthermore, the lab often does not have information regarding a person’s race. Therefore, most labs report 2 numbers: 1 for African Americans and 1 for non-African Americans. Finally, the MDRD equation has not yet been validated for persons over 70 years old, and therefore many labs do not report any eGFR results for older individuals.
At the Mayo Clinic our renal function lab can directly measure GFR using iothalamate clearance. This slide depicts measured GFR in 3 large groups of patients tested in our lab: Normal individuals being evaluated as potential kidney donors, persons that have already donated a kidney, and persons with known chronic kidney disease. The slide shows measured GFR, or iothalamate clearance, on the Y-axis, and the MDRD estimated GFR on the X-axis. Persons with chronic kidney disease are shown in the small black dots. The black line shows the correlation of measured GFR and estimated GFR, and falls directly over the line of identity. Therefore, the MDRD equation estimates GFR very well in patients with known chronic kidney disease. The normal potential kidney donors are shown in blue, and have measures GFR that is slightly higher than the estimated GFR. This data shows that the MDRD equation underestimates GFR in persons with normal kidney function. Finally, the red circles and line show data from persons post-kidney donation. Their values fall roughly halfway between the chronic kidney disease and normal populations.
Although creatinine is very useful marker of GFR, given the problems discussed above people have looked for alternatives. One example is cystatin C, a low-molecular-weight protein that is also freely filtered in the kidney. Compared to creatinine, production of cystatin C is much less influenced by a person’s age, gender, and size. Therefore, blood levels of cystatin C better reflect GFR than serum creatinine, and over recent years much effort has been spent evaluating cystatin C as a GFR marker.
The Mayo Clinic renal laboratory has recently evaluated a new cystatin C assay. It is a particle-enhanced turbidometric immunoassay, or PETIA, and can be run on a chemistry autoanalyzer. This confers certain lab advantages including high-capacity, quick turnaround time, and lower-cost compared to other platforms. Importantly, this assay is also standardized to a cystatin C international reference material. Our validation studies revealed very low within and between run coefficient of variations of 0.65-1.33%. The lower limit of quantification was 0.35 mg per liter, with an upper range of 6 mg per liter.
Previously, the Mayo Clinic renal laboratory used a nephelometric assay for cystatin C. When samples were recently run using both platforms, we found that there was a 23% bias, with the newer PETIA assay being higher than the previous nepholmetric assay or PENIA.
Therefore, we pulled some older samples that had been run on the PENIA in 2000 and stored at -70 degrees since. Surprisingly, they were 19% lower when rerun on the PENIA in 2010. Therefore, there appears to have been assay drift in the commonly used PENIA platform between 2000 and 2010. Importantly, recent publications have reported a similar shift in values using the same nephelometric assay for cystatin C in other laboratories. The standardization of the new turbidometric assay to an international cystatin C reference material is a major advantage, and should prevent similar drift in values over time.
Since cystatin C production is not greatly influenced by patient demographics, equations to estimate GFR using cystatin C are simpler. Shown on this slide is 1 equation developed using the PETIA assay. Note that there are no variables for age, race, body size, or gender. Recently we evaluated 107 persons in the renal function lab who are having clinically indicated GFR measurements by iothalamate clearance. We also measured blood cystatin C and creatinine to calculate their estimated GFR. Shown here is the good comparison between measured GFR and cystatin C estimated GFR. In general, there is a very good correlation, without significant bias across the GFR range.
We also calculated MDRD estimated GFR from serum creatinine in the same population. In general, both equations performed reasonably well. However cystatin C. estimated GFR was more specific for identifying patients with a borderline reduction in GFR between 60 and 80 ml/min/1.73m2, shown in the highlighted region on the table. Therefore, cystatin C may have particular utility for providing accurate estimates of GFR among patients with stages 1 and 2 CKD, and for not misdiagnosing normal patients with CKD.
We have also established a reference range for cystatin C. As shown here, values did trend higher for persons over 60, although this was not statistically significant in the number of individuals studied to date.
Currently, our reference ranges for infants and children up to 17 years of age are based on literature values, as shown on the slide. The adult values were obtained from the reference range study shown the previous slide. Our current 95% reference range is 0.488 to 1.134 mg/L, with a mean of 0.71 mg/L. There were no gender differences apparent in our study. However, we are currently evaluating additional samples to determine if the reference range should be slightly higher for those greater than 50 years of age. Of note, other large epidemiological studies have suggested that cystatin C levels do increase in older individuals.
Several large studies have recently suggested a particularly useful niche for cystatin C testing. In the next 2 slides I will show some data from 2 of these studies. The REGARDS study included over 26,000 adults monitored over 7 years for risk of stroke. Data were also analyzed for other outcomes including risk of CKD. As shown in the table on the left, an elevated serum creatinine alone at baseline did identify persons at risk for subsequent CKD. Cystatin C alone was slightly better, while an elevated albumin creatinine ratio in the urine, or ACR, was better than either creatinine or cystatin C alone. Persons with an increased cystatin C and urinary albumin creatinine ratio at baseline had much higher rates of subsequent CKD. This combination performed better than creatinine and urine albumin creatinine ratio for predicting risks. Persons with an elevated creatinine and cystatin C also had much higher rates of CKD than either marker alone. Finally, combining all three markers clearly identified patients at very high risk for subsequent CKD.
The table on the right shows that risk of mortality was also associated with cystatin C levels. Persons with a normal eGFR by MDRD using serum creatinine, but an increased cystatin C level, had a 2-fold increased risk of mortality over the 7 years of the study. This increased to 3-fold if urinary albumin creatinine ratio was also increased. Mortality was highest in persons and had an increased creatinine, cystatin C, and urine albumin creatinine ratio, 5.6-fold above those with kidney function that was normal by all criteria.
This slide shows data from close to 12,000 other participants in 2 studies designed to assess cardiovascular outcomes over time. Once again, the investigators also looked at predictors of kidney function in their cohorts. In this study an elevated cystatin C predicted all cause mortality and cardiovascular disease, while creatinine alone did not. Furthermore, an increase in cystatin C was a much stronger predictor of future kidney failure than creatinine alone. An increase in creatinine together with cystatin C markedly increased the risk of kidney failure, close to 24 times baseline. Taken together, both of these large studies suggest that cystatin C is particularly good at identifying persons at risk for progression of chronic kidney disease, as well as future cardiovascular events and even mortality.
Another setting where cystatin C may have particular advantages over creatinine is in the acute hospitalized patient. Creatinine is produced in muscles, and muscle mass declines quickly with acute illness. Furthermore, blood levels of creatinine often decline acutely in the hospital due to dilutional effects of intravenous fluids. In this prospective study involving intensive-care-unit patients, cystatin C often increased before creatinine, at a threshold of either 50% or 25% over baseline. The average lead time for cystatin C elevation over creatinine was 4 to 6 hours, although larger in many individual patients. Therefore, cystatin C may be useful for detecting acute kidney injury earlier among hospitalized patients. It may also reflect overall GFR better in these acutely ill hospitalized patients, even without acute kidney injury, and help physicians initiate appropriate treatments earlier, and most importantly dose renally excreted drugs more accurately. This is a promising area for use of cystatin C testing that will require further and rigorous evaluation.
Now let us go back to our patient we started off the presentation discussing. As you recall, he is a 66-year-old African American gentleman with a serum creatinine of 1.5 mg/dL. His MDRD estimated GFR comes out to 57 mL per minute per 1.73 m2. Using the CKD-EPI equation we get a very similar number of 55 mL per minute per 1.73 m2. Therefore he clearly has stage 3 chronic kidney disease, and is potentially at risk for further loss of kidney function.
At this point, it would be extremely helpful to further understand his risk, especially since this was 1 of his key questions. Two tests that are extremely helpful to address this question are the urine albumin creatinine ratio and a serum cystatin C. In the table on this slide, I have used the data from the 2 large studies previously discussed in order to roughly estimate what his risk of end-stage renal disease would be based upon hypothetical results of the these 2 additional tests. The urine albumin creatinine ratio could either be normal at 20 mg per gram, or slightly high at 100 mg per gram. Similarly his cystatin C could either be normal at 1.1 mg/L which equates to an estimated GFR of 68 mL per minute per 1.73 m2 or it could be slightly elevated at 1.4 mg/L to give an estimated GFR of 47 ml/min/1.73m2. If he had normal cystatin C and normal albumin creatinine ratio his risk of end-stage renal failure is 4 times higher than persons without an elevated creatinine at all. However, if his cystatin C is also high this risk increases to 31 times above baseline. Furthermore, if cystatin C and albumin creatinine ratio are both high his risk increases to over 400 times baseline values. These calculations provide an idea how use of cystatin C could help us to identify patients at particularly high risk of future renal failure and cardiac vascular events.
Potential interventions for our patient would include careful control of blood pressure with a target systolic value less than 130 mmHg. If he had proteinuria, one would consider use of specific blood pressure agent such as ACE inhibitors or angiotensin receptor blockers. If he fell into the high risk group with an elevated cystatin C, one might consider more aggressive targets for serum lipids. Other lifestyle intervention such as weight loss and quitting smoking would also be advisable. Identifying him at very high risk for complications might make him more amenable to such lifestyle alterations. Finally, careful control of diabetes would also be indicated in such an individual.
In conclusion, blood levels of cystatin C reflect GFR better than blood levels of creatinine alone. However, equations that use creatinine and demographics to predict an estimated GFR perform about as well as cystatin C alone. Elevated cystatin C levels identify patients at greater risk of CKD progression and early death, largely from cardiovascular morbidity. And finally, in the hospital setting cystatin C levels may be a better indicator of GFR, and levels of cystatin C may increase earlier than creatinine among patients that experience acute kidney injury. Thank you.