Biomarkers of Acute Renal Failure
Click CC to turn on closed captioning.
Published: August 2009Print Record of Viewing
The use of various biomarkers for the identification of acute renal failure and, specifically, the use of neutrophil gelatinase-associated lipocalin (NGAL).
Presenter: John C. Lieske, MD
- Division of Clinical Core Laboratory Services at Mayo Clinic
Questions and Feedback
Welcome to Mayo Medical Laboratories' Hot Topics. These presentations provide short discussion of current topics and may be helpful to you in your practice. You may access a handout of this presentation to make notes while viewing this video hot topic. Closed Captioning also is available by selecting the CC box in the lower right corner of the presentation. At the end of this presentation, you will be taken to a brief survey. We appreciate any comments or suggestions of future Hot Topics that you may have.
Our speaker for this program is John C. Lieske, MD, Professor of Medicine, Medical Director of the Renal Function Laboratory, and Consultant in the Central Clinical Laboratory in the Department of Laboratory Medicine & Pathology at Mayo Clinic in Rochester, Minnesota. Dr. Lieske will be discussing the use of various biomarkers for the identification of acute renal failure and, specifically, the use of neutrophil gelatinase-associated lipocalin (NGAL).
Acute Kidney Injury
The topic for today is acute kidney injury which is defined as a sudden decrease in kidney function. Previously this was often called acute renal failure. If a patient has chronic kidney disease, but suffers an acute decline, this is commonly called acute on chronic renal failure. Acute kidney injury is often reversible, although not always, and is largely a disease of acutely ill, hospitalized patients. Acute kidney injury carries high morbidity and mortality. Affected patients are at increased risk of infection, bleeding, and cardiac events. Currently, treatment is largely supportive for the patient until the kidneys can recover. This supportive care often includes dialysis. Because of this need for prolonged hospitalization and intensive treatments such as dialysis, acute kidney injury is very costly to the health care system. Therefore, efforts to improve early diagnosis and prevention of acute kidney injury are of paramount importance.
Common Causes of Acute
Acute kidney injury can be caused by anything which damages the kidneys. Renal ischemia, or decreased blood flow to the kidneys, is a very common cause. Cardiac disease, hemorrhage, or septic shock are common examples of this. Certain toxins can adversely affect renal cells. Aminoglycoside antibiotics and IV contrast dye are 2 examples of that. Sometimes allergic reactions to drugs are localized to the kidney, causing interstitial nephritis. Anything that causes obstruction to the urinary outflow can lead to an acute decline of kidney function. Examples might include prostatic obstruction or bladder cancer. Finally, certain vasculitides can affect the kidney and lead to relatively rapid loss of kidney function. Often times, many of these factors are occurring at the same time and, therefore, many causes of acute kidney injury are multifactorial. For example, many patients might have septic shock as well as exposure to nephrotoxic antibiotics or contrast dye.
However, of the potential causes, renal ischemia due to decreased cardiac function and/or septic shock is probably the most common underlying theme amongst hospitalized patients with acute kidney injury.
Progression From Pre-renal to Acute Tubular Necrosis
Typically patients with acute kidney injury progress from a period of prerenal decline in kidney function, where they are at risk for structural kidney damage, as shown on this slide. These patients are in a prerenal state due to some initial insult that decreases blood pressure and renal perfusion. At some point the situation is compounded by a second insult, for example, infection or administration of a nephrotoxic agent, or perhaps the patient has prolonged and severe hypotension such that irreversible ischemic damage occurs to the kidney. This is indicated by the second insult on the slide. At this point, the patient is destined to develop true acute tubular necrosis. Nevertheless, serum creatinine is not likely to begin to rise for another 24 to 48 hours. During this time, progressive kidney damage is continuing to occur. It would be ideal to identify patients very early after injury so that appropriate interventions can take place. This is indicated by the “window of opportunity” on the slide, which spans the prerenal stage of acute kidney injury through early acute tubular necrosis. This would, in many ways, parallel the situation in acute myocardial infarction. There are patients identified very early with elevated enzymes such as troponin. Such patients can be effectively treated with thrombolytics and other interventions in order to save additional myocardium that is at risk. In a similar manner, if we can identify patients with acute kidney injury very early, then we could possibly save them from additional kidney damage and impact on their overall course. Examples of potential interventions would include specific medications or growth factors that might hasten kidney recovery or limit damage. In fact, previous studies using potentially promising growth factors may have failed to demonstrate benefit because patients were already at the stage where their creatinine had gone up and a large amount of irreversible kidney damage had already occurred.
Towards a Kidney Troponin
The biomarker field is fairly well developed in cardiology. Release of enzymes such as AST, CPK, and LDH were identified in the 1950s and 60s. These have been refined to include CPK MB and then troponin over the last several decades. Currently, additional biomarkers are being developed that may better identify subtle ongoing ischemia within the myocardium. In comparison, in nephrology, serum creatinine has remained the only real biomarker for many decades. To be fair, there are promising new biomarkers on the horizon, many of which I will discuss today, so this slide may need to be updated soon. The point remains, however, that we have a long way to go catch up to the cardiologists.
Origin of Formed Elements in Urinalysis
It is not quite fair to say that creatinine is the only biomarker of acute kidney injury. Urinalysis is a very helpful test for identifying evidence of kidney damage. Ordinarily there are very few formed elements in the urine. However, after acute kidney injury, renal epithelial cells can be shed into the tubular lumen and detected in the urine. Furthermore, these can form renal epithelial cell casts when they gel with Tamm Horsfall protein that is present in everyone’s urine. Other elements in the urinary tract can also form casts, including white blood cells that might indicate infection and red blood cells, which typically indicate glomerular inflammation.
Urinanalysis: A Traditional Biomarker of AKI
This table indicates how the urinalysis can be used to help categorize patients that have evidence for acute kidney injury, by which we usually mean an elevated serum creatinine. If the decrease in renal function is entirely prerenal in nature and due to decreased renal blood flow, then we would expect to find no formed elements in the urine. In acute tubular necrosis, one can find renal epithelial cells, granular casts, and renal epithelial cell casts. For infection, one might expect to find white cells and white cell casts. Patients with nephrotic renal disease will have evidence for lipid casts and oval fat bodies, while patients with glomerulonephritis often will have the red cells, red blood cell casts, and dysmorphic red cells. The next few slides will show some images to demonstrate what these look like.
Hyaline casts are formed by Tamm Horsfall protein, which has precipitated inside of individual tubule lumens. Tamm Horsfall protein is, in fact, the most common protein in the normal human urine, and is secreted by cells in the thick ascending limb. It has a peculiar characteristic of aggregating to form a gel when it is concentrated in urine. If there’s nothing else around when it gels, it forms these hyaline casts, which are normal and only indicate that the patient may have had somewhat concentrated urine at the time it was collected. If there are other cells or structures in the urine at the time Tamm Horsfall protein precipitates, they will be trapped. In fact, Tamm Horsfall protein is the glue that holds casts together.
In this slide we see an example of a renal epithelial cell cast. The renal epithelial cells can be recognized by their round distinct nucleus. One can appreciate how they are trapped by the Tamm Horsfall protein and held together to form this cast. Renal epithelial cells can also be detected free in the urine, and would also indicate ongoing acute kidney injury.
This slide shows an example of a white blood cell or leukocyte cast. These are typically seen in ongoing kidney infection or pyelonephritis. They could also be seen in acute interstitial nephritis, in which case eosinophils might also be found in the cast.
This is an example of a red blood cell cast. The red cells can be differentiated by their clear outline, which is most clearly seen in the middle of the cast. A red cell cast should indicate ongoing glomerulonephritis. In this circumstance, one might also expect to find free red cells in the urine. Sometimes these red cells become damaged while being squeezed through glomerular holes and end up dysmorphic or misshapen. In addition to red cells and red cell casts, one would expect to find abnormal amounts of urinary protein and albumin in patients with glomerulonephritis.
The renal epithelial cells that makeup a renal epithelial cell cast can degrade while they are present free in the tubular lumen. In this case, they can form a pigmented or granulated cast. These are less specific than a true renal epithelial cell cast, since other structures such as red cells or white cells can also degrade and form granular casts. Usually, however, large numbers of pigmented granular casts, so called muddy brown casts, are indicative of ongoing acute tubular necrosis (ATN).
This slide shows an example of a granular cast, again most commonly associated with ATN.
Other Helpful Indices
In addition to the urinalysis, urine electrolytes can also be helpful for differentiating prerenal causes of decreased renal function from true acute kidney injury. If a patient has oliguria or decreased urine output due to a prerenal cause, one would expect the kidney to avidly reabsorb sodium and chloride. Therefore, the concentration of urinary sodium or chloride should be reduced to <10 mEq/L. This can be normalized to creatinine by calculating a fractional excretion of sodium or chloride. Thus, a fractional excretion of sodium <1% would be consistent with a prerenal cause. Looked at another way, if a person’s kidney is able to reabsorb sodium effectively and achieve a fractional excretion of sodium <1%, this is a sign that true cellular injury or ATN has not yet occurred. Many patients in the hospital have received diuretics, which would tend to increase urinary sodium and chloride excretion. In this circumstance, calculating a fractional excretion of urea may also be useful, since urea handling by the kidney is less influenced by diuretics than sodium and chloride. A fractional excretion of urea <35% is more consistent with prerenal causes of decreased kidney function.
Potential Sources of Urinary Biomarkers
In order talk about potential urinary biomarkers of kidney injury, it is first helpful to consider what their source might be. The glomerulus is usually a fairly effective barrier to filtration of proteins from the blood as long as they are >40 to 50 kDa in size. Therefore, much less than 1% of serum albumin ordinarily would make it past this filtration barrier into the urine. However, some albumin does make it through this filter. In addition, smaller molecular weight proteins are freely filtered. All proteins that are filtered into tubular fluid are reabsorbed in the proximal tubule and processed within lysosomes, as shown on the slide. Under conditions of proximal tubular injury, one might expect to find increased levels of these smaller molecular weight proteins, as well as some albumin. Cystatin C is another example of a small molecular weight protein that is metabolized in the kidneys by this mechanism, and, in fact, urinary cystatin C has been proposed as a sensitive biomarker of acute kidney injury. Other proteins are directly released in the urine by injured kidney epithelial cells. Examples include brush border enzymes such as NAG and GGT, as well as proteins that are upregulated in injured cells like NGAL and KIM 1. Finally, acute tubular necrosis is accompanied by some degree of inflammatory response in the region of cell damage and inflammatory markers such as IL18 have been identified in the urine of patients with acute kidney injury.
NGAL as a Biomarker
Neutrophil gelatinase-associated lipocalin, or NGAL, has been recently identified as a particularly promising potential biomarker of acute kidney injury. This 25-kDa protein was initially described in neutrophils, as its name implies. However, it is also expressed in low amounts in many human tissues, including kidney cells. Importantly, expression of NGAL is induced in injured tissues. This up regulation of NGAL expression is quite rapid, usually within 2 to 4 hours of injury. Fortunately, NGAL is also very stable and easily detected in urine.
Studies Support NGAL as a Biomarker of AKI in the Following Situations
Over the last several years, many clinical studies have been completed looking at urinary NGAL as a biomarker of acute kidney injury. It appears to be a very rapid and useful predictor of acute kidney injury amongst pediatric patients having cardiac surgery or at risk for contrast-induced kidney damage. A recent study in adults also found that increased urinary NGAL release was a good predictor of subsequent kidney injury in patients admitted from a large emergency room. Unfortunately, NGAL also increases somewhat in the urine of patients with chronic kidney disease. Therefore, it may be that NGAL is a better marker of acute injury in pediatric patients as opposed to adult patients, since chronic kidney disease is more commonly a disease of adults. Already mentioned, NGAL is very stable in urine, which does make it a very convenient analyte for the laboratory.
Mayo Renal Laboratory Urinary NGAL Normals
We have recently validated a urinary test for NGAL in the Mayo Clinic Renal Function Laboratory. As part of this effort, we have developed a reference range for NGAL amongst a group of normal, nonhospitalized individuals. Cutoffs for normal would be approximately 65 ng/mL, or 81 nanograms/gram creatinine.
Normal Urinary NGAL Levels are Not Dependent on Age, but are Higher in Women than Men
When we looked at this normal value data, we found that urinary levels of NGAL were not dependent on age. Urinary NGAL was, however, higher women than in men. Therefore, we will use gender-specific reference ranges. The reason for this gender difference in baseline urinary NGAL excretion is not currently known.
Ongoing Mayo Clinic Clinical Validation Study
We are currently completing a clinical validation study at Mayo Clinic, in order to determine optimal cutoff ranges to differentiate normal and disease. To do this, we are looking at urine samples from patients admitted from our emergency room. We will obtain traditional urinalysis, urinary electrolytes, and urinary NGAL levels. We will then determine clinical outcomes by retrospective review of the medical record. Key outcomes will be acute kidney injury by the AKIN criteria as well as mortality. The AKIN criteria were developed by the Acute Kidney Injury Network, and grade severity of acute kidney injury in stages 1-3 based upon degrees of elevation in creatinine and decreases in urinary output.
Ongoing Mayo Clinic Clinical Validation Study: Preliminary Results
This study is ongoing, and we hope to have the data analyzed soon. However, preliminary results indicate that the presence of renal epithelial cells or granular casts in the urinalysis is still a very specific marker of acute kidney injury. However, urinary NGAL is somewhat more sensitive than this traditional marker. We plan to use the complete data set in order to construct receiver operating curves and develop appropriate cutoff values to differentiate cases of acute kidney injury with maximum sensitivity and specificity.
Kidney injury molecule 1, or KIM-1, is another promising urinary biomarker of acute kidney injury. Like NGAL, it was identified first in animal models, looking at genes and proteins that were up regulated after injury. Expression of KIM-1 is increased in renal cells 12 to 24 hours after injury. This increased KIM-1 expression is mostly by proximal tubular cells, and since the time course of the increase is a little later than NGAL, both tests may turn out to be complementary. For example, NGAL may be quite useful for early detection of acute kidney injury, while KIM-1 picks up cases a little later at time points when NGAL expression may have peaked and started to go down again.
Other Potential Markers on the Horizon
Other potential urinary biomarkers are being actively studied in clinical trials. Examples include liver fatty acid binding protein (L-FABP), urinary cystatin C, and urinary interleukin-18. The time course of expression, sensitivity, and specificity of each remain to be determined.
Mayo Clinic Goals
Currently, the Mayo Clinic Renal Function Laboratory is actively developing methodology for monitoring these and other urinary biomarkers of kidney injury. We have validated a rapid ELISA test for NGAL. Turnaround time for this is approximately 1 hour. As the literature evolves, we plan to develop additional biomarkers. A potential strategy might be to develop a multiplexed panel of acute kidney injury biomarkers that could be used to stage patients at risk. It’s also possible that many of these markers will have utility for following patients with chronic kidney disease in order to gauge the effectiveness of ongoing therapies to slow progression.
In conclusion, acute kidney injury is a significant public health problem with large morbidity and mortality, as well as health care cost. The current clinical tests such as serum creatinine do not allow early detection of cases, at a time when interventions might have the biggest effect to prevent the development of full-blown acute kidney injury. For this purpose, urinary biomarkers under development show great promise. Of these, NGAL has the most extensive literature to date. However, it is likely that other biomarkers will also prove useful in this rapidly evolving area. Longer term, it will be important to perform clinical outcome studies using the most useful biomarkers. Only then will we be able to prove the hypothesis that early detection of kidney injury and rapid initiation of therapy actually will improve patient outcomes, which is obviously our primary objective.