Review: Serum and urine biomarkers of kidney disease: A pathophysiological perspective


Dr Greg H Tesch., Department of Nephrology, Monash Medical Centre, 246 Clayton Road, Clayton, Vic. 3168, Australia. Email:


The use of reliable biomarkers is becoming increasingly important for the improved management of patients with acute and chronic kidney diseases. Recent developments have identified a number of novel biomarkers in serum or urine that can determine the potential risk of kidney damage, distinguish different types of renal injury, predict the progression of disease and have the potential to assess the efficacy of therapeutic intervention. Some of these biomarkers can be used independently while others are more beneficial when used in combination with knowledge of other clinical risk factors. Advances in gene expression analysis, chromatography, mass spectrometry and the development of sensitive enzyme-linked immunosorbent assays have facilitated accurate quantification of many biomarkers. This review primarily focuses on describing new and established biomarkers, which identify and measure the various pathophysiological processes that promote kidney disease. It provides an overview of some of the different classes of renal biomarkers that can be assessed in serum/plasma and urine, including markers of renal function, oxidative stress, structural and cellular injury, immune responses and fibrosis. However, it does not explore the current status of these biomarkers in terms of their clinical validation.


Kidney damage can be caused by a wide range of insults including infections, toxins, ischaemia, hypertension, genetic or metabolic disorders, autoimmune diseases or allograft rejection. The effects of these insults may induce acute kidney injury, which is clinically defined as a sudden reduction in renal function or urine output,1 or they may promote the development of chronic kidney disease (CKD), in which kidney structural or functional alterations persist for at least 3 months.2 Determining the nature and severity of this injury as early as possible is a prime goal for therapeutic intervention and successful patient management. Biological markers (biomarkers), which identify normal or pathogenic processes, or responses to treatment, are a valuable tool for determining a patient's condition. Biomarkers can be used to assess a predisposition towards an illness or detect biological abnormalities, but are more often used to diagnose and measure a pathological condition or make a prognosis about the development of disease. They can also be useful for evaluating the response to a particular therapy. Biomarkers do not need to be involved in the disease process and in this respect are different to risk factors such as age, obesity and smoking, which are associated with a disease because they play a role in causing it.


The characteristics of a biomarker need to be carefully considered before its potential usefulness can be determined. Some important criteria for selecting renal biomarkers are listed in Table 1.

Table 1.  Preferred characteristics of a renal biomarker
1) It can be measured easily, accurately and reproducibly
2) It can sensitively indicate renal injury or the kidney response to treatment
3) It provides useful and cost-effective clinical information that is easy to interpret
4) It gives information that is additive to that of conventional clinical factors
5) It can identify or differentiate specific types of renal injury or kidney disease
6) It is applicable across a variety of populations (race, gender and age)

Ideally, these biomarkers should be obtainable by procedures that are either non-invasive (e.g. urine collection) or have minimal effects on patients (e.g. routine blood collections). Consequently, large efforts have been made to identify reliable biomarkers of renal injury in serum, plasma and urine.


Recent technological advances have resulted in the identification of a growing number of potential renal biomarkers in the serum and urine of patients and animal models of kidney disease. Many of these are still awaiting further testing and clinical validation. However, it is becoming clear that these renal biomarkers can be grouped into different categories (Table 2), which represent different types of renal injury. These categories are discussed individually below.

Table 2.  Renal biomarker categories
ClinicalFunctionalOxidative StressStructural & Cellular InjuryImmune ResponsesFibrosis
Blood PressureGlomerular Filtration RateAdvanced Glycation End ProductsAlbuminuriaChemokinesProfibrotic Growth Factors
Urine OutputPodocyte Injury MarkersInflammatory Cytokines
Cell Adhesion Molecules
Genetic ScreeningCapillary Pressure MarkersAdvanced Lipoxidation End ProductsTubular Injury markersImmunoglobulinSoluble Matrix Fragments
Capillary Flow MarkersReactive Oxygen SpeciesComplement Components
Reactive Nitrogen Species 


Blood urea nitrogen (BUN) and creatinine clearance are well-established biomarkers of renal function that can be measured cheaply and easily. Both urea and creatinine are products of protein metabolism, which are cleared almost entirely by the kidneys. BUN is routinely measured in serum by an enzyme/oxidation reaction assay; however, its levels are affected by non-renal influences such as protein intake, dehydration, liver function, gastrointestinal bleeding and steroid use.3 In addition, BUN assays often underestimate renal function due to interfering chromogens. Creatinine levels in serum and urine can be measured by a variety of assays (Jaffe rate reaction, creatininase method, high-performance liquid chromatography (HPLC) method), but are most commonly assessed by the Jaffe rate reaction, which is cheap and easy to perform. However, HPLC is the most sensitive method for assessing creatinine levels and is not affected by chromogen interference.4 Creatinine levels are also affected by non-renal influences such as muscle mass, age, gender and liver function.5 Creatinine clearance is one of the most common assessments of renal function but it lacks sensitivity when renal impairment is mild and can be affected by tubular secretion of creatinine when the glomerular filtration rate is declining.

Cystatin-C has recently emerged as a reliable alternative biomarker of renal function. Cystatin-C is a cysteine protease inhibitor that is constantly produced by nucleated cells and released into the blood, where it is normally reabsorbed and catabolized by kidney tubules without re-entering the blood stream.6 Serum levels of cystatin-C can currently be measured by immunonephelometry or enzyme-linked immunosorbent assay (ELISA) and are affected by steroid use or thyroid dysfunction.6 Cystatin-C is particularly sensitive at detecting changes in kidney function when renal impairment is mild,7 and is better than creatinine for assessment of acute kidney injury due to its shorter half-life.8

Some other potential biomarkers of renal function are also worthy of note. Uric acid is normally excreted through the kidney but circulating levels increase during renal impairment in CKD. Animal model studies have shown that hyperuricaemia activates the renin-angiotensin system, induces oxidative stress and reduces renal function.9 Increased levels of serum uric acid have been detected in patients with CKD by colorimetric assay and predict a greater risk of end-stage renal disease.9 Urinary levels of angiotensinogen detected by ELISA have been reported to be a specific index of the intrarenal renin-angiotensin system and correlate with blood pressure and glomerular filtration rate in CKD.10 Therefore, urine angiotensinogen appears to be a potential biomarker of renal function in kidney diseases that are dependent on hypertension. The fractional excretion of magnesium (FE Mg) is considered to be a measure of tubular function because tubules normally reabsorb magnesium filtered by glomeruli.11 Levels of magnesium can be measured in serum and urine by atomic absorption spectroscopy. Elevations in the FE Mg are thought to indicate the loss of peritubular capillary flow resulting from tubulointerstitial damage.11


Oxidative stress is known to play a pathological role in animal models of CKD.12 Increased oxidative stress is also present in patients with moderate to severe CKD;13 however, further longitudinal and intervention studies are required to help define the role of oxidative stress in the development of human CKD. Some serum and urine biomarkers have been shown to reliably measure the level of renal oxidative stress in patients and animal models. During oxidative stress, oxidized guanine in cellular DNA is spliced out by DNA repair enzymes, releasing a metabolically stable product 8-hydroxy-2-deoxyguanosine (8-OH-dG) into the urine. Increased levels of 8-OH-dG can be detected in urine by ELISA during CKD.14 Peroxidation of lipids also occurs during oxidative stress, resulting in the formation of 8(F2a)-isoprostane and 4-hydroxy-2-nonenal. Levels of 8-isoprostane and 4-hydroxy-2-nonenal can be measured in serum or urine by ELISA or HPLC and are elevated in CKD.15–17 In addition, renal oxidative stress produces peroxynitrite that nitrates protein tyrosine residues to form stable 3-nitrotyrosine peptides. A recent study has indicated that levels of 3-nitrotyrosine peptides can now be accurately measured in serum or urine using liquid chromatography and mass spectroscopy, which may prove to be useful for assessing both oxidative and nitrosative stress in kidney disease.18

Advanced glycation end products (AGE) are proteins modified by oxidative stress that accumulate in the blood of diabetic and uremic patients. These circulating AGE can deposit in the kidney and cause cellular dysfunction and renal damage. Elevated serum and urine levels of the AGE pentosidine can be detected by HPLC or ELISA and help to predict the development of diabetic nephropathy.17 In addition, plasma levels of pentosidine have been shown to increase with loss of residual renal function in patients on peritoneal dialysis and to decrease with patients recovering renal function after transplantation.19,20


The excretion rate of albumin is the most commonly used biomarker of renal injury. Albumin is the most abundant protein in the circulation and during normal kidney function very little intact albumin is excreted by the kidney (<30 mg/day in humans). However, following renal injury, glomerular filtration of albumin is increased and the reabsorption and degradation of albumin by tubules are decreased, resulting in increased levels of intact albumin in the urine (i.e. albuminuria). Patient albuminuria is usually classified by ranges of severity, which are: microalbuminuria (30–300 mg/day), macroalbuminuria (300 mg–3 g/day) and nephritic range albuminuria (>3 g/day). Albuminuria is commonly used as an early marker of renal injury because it often precedes a decline in renal function. However, it cannot distinguish different types of proteinuric kidney disease and has a limited ability to predict disease progression and determine therapeutic efficacy. Albuminuria is commonly measured by immunological techniques, which include: immunonephelometry, immunoturbidimetry, radioimmunoassay and ELISA.21 These techniques are good for assessing albumin excretion, which is distinctly higher than normal. However, newer HPLC-based methods (e.g. the Accumin Test) can identify both immunoreactive and non-immunoreactive albumin providing greater sensitivity than conventional immunological methods for distinguishing microalbuminuria from normal albumin excretion.22,23

Podocyte injury is a feature of many kidney diseases that is postulated to increase glomerular filtration of albumin. Severely damaged podocytes can detach from the glomerular basement membrane and be collected in the urine sediment. Analysis of the urine sediment by quantitative PCR or ELISA can determine mRNA or protein levels of podocyte-specific molecules (e.g. nephrin, podocin, podocalyxin) as markers of podocyte injury. Increased urine sediment levels of nephrin and podocin have been detected in patients with diabetic nephropathy and active lupus nephritis.24,25 Similarly, increased levels of podocalyxin have been found in the urine sediment of patients with IgA nephropathy, lupus nephritis and post-streptococcal glomerulonephritis.26

Sensitive markers of tubular injury have been identified in acute and CKD. N-acetyl-beta-D-glucosaminidase is a proximal tubular lysosomal enzyme, which is released during damage to proximal tubules. Increased urine levels of N-acetyl-beta-D-glucosaminidase can be detected by enzymatic assay in kidney diseases that involve tubulointerstitial damage.27,28 Kidney injury molecule-1 is a transmembrane protein that is expressed on the luminal surface of proximal tubules during injury. Increased urine levels of kidney injury molecule-1 can be detected by ELISA, microbead assay or immunochromatographic dipstick in patients with tubulointerstitial damage and correlate with renal expression.28–30 Liver-type fatty acid-binding protein (L-FABP) is a marker that is shed by proximal tubular cells in response to hypoxia from decreased peritubular capillary flow. Urine levels of L-FABP are a sensitive indicator of acute and chronic tubulointerstitial injury.31,32 In CKD, increasing urine levels of L-FABP correlate with declining renal function.32 L-FABP is not assessable in kidney disease models that use C57BL/6 mice, because these mice have a regulatory defect that suppresses L-FABP expression.33 Neutrophil gelatinase-associated lipocalin (NGAL), also known as lipocalin-2, is an iron-transporting protein that is almost entirely reabsorbed by tubules in the normal kidney. NGAL levels in the urine increase following acute nephrotoxic and ischaemic insults, indicating defects in proximal tubular reabsorption and the distal nephron.34 Urine levels of NGAL can be measured by ELISA and are a very sensitive marker of acute kidney injury, which can increase up to 1000-fold in patients.35 Urinary NGAL has also been used as a triaging tool to randomize patients with AKI to treatment.36 In addition, serum and urine NGAL levels have been found to be independent risk markers of CKD.37

Recent research has indicated that levels of exosomal transcription factors may also be used to identify kidney injury. Exosomes are tiny vesicles that are excreted by epithelial cells in normal and diseased kidneys. These exosomes contain transcription factors that can be activated by pathological stimuli. Exosomes can be collected from fresh or frozen urine by ultracentrifugation and have been assessed for transcription factors by western blotting. Urine exosomal levels of ATF3 are increased during acute but not CKD.38 In contrast, exosomal levels of podocyte WT-1 are increased during focal segmental glomerulosclerosis (FSGS) and precede albuminuria, but are not elevated in acute kidney injury.38


Molecular components of humoral immunity (e.g. immunoglobulin, complement components) and cellular immunity (e.g. chemokines, leukocyte adhesion molecules, pro-inflammatory cytokines and their soluble receptors) are known to play significant roles in the development of renal inflammation. The serum or urine levels of these molecules can be detected by ELISA and some have been shown to be sensitive markers of the immune response in the injured kidney.

Urine excretion of immunoglobulins can predict the development of immune-mediated kidney diseases. The fractional excretion of IgG has been shown to predict the progression of primary FSGS and the response of this disease to treatment.39 Similarly, urine levels of IgA can be an indicator of the severity of renal damage in IgA nephropathy and are known to correlate with proteinuria, serum creatinine and glomerulosclerosis in this disease.40 In comparison, urine levels of IgM are a strong predictor of disease progression for patients with anti-nuclear cytoplasmic antibody-associated vasculitis.41 Furthermore, because IgM has a high molecular weight (600 kDa) and is usually not filtered by healthy glomeruli; its levels in urine are a stronger predictor of end stage renal disease than the more readily filtered albumin (68 kDa) in a number of glomerular diseases.42 However, these filtration properties of IgM suggest that it is better associated with advanced glomerular injury and is not a specific or sensitive marker of early renal damage.

Levels of complement C3d, C4d and complement factor H have been identified as potential biomarkers of complement-mediated injury in renal diseases. Increased urine levels of C3d are found in tubulointerstitial nephritis, membranous nephropathy and non-membraneous glomerular diseases.43 In patients with glomerular diseases, the urine excretion of C3d correlates with the progression or remission of proteinuria and is independent of the underlying glomerular disease.43 A study has also shown that serum C4d and urine C3d correlate with moderate to severe disease activity in lupus nephritis.44 In addition, urine levels of factor H (a regulator of the alternative pathway of complement) are elevated in patients with IgA nephropathy and idiopathic membranous nephropathy and are associated with disease activity.45,46

During a renal inflammatory response, leukocytes are recruited into the kidney by chemokines. The urine levels of some chemokines increase with the development of renal inflammation and correlate with kidney leukocyte numbers. Monocyte chemoattractant protein-1 (MCP-1), also known as CC-chemokine ligand 2, is considered to be the most potent chemokine for recruiting monocyte/macrophages. It is expressed by many cell types in diseased kidneys, but is produced mostly by glomerular and tubular epithelial cells.47 Urine levels of MCP-1 correlate with kidney MCP-1 expression and interstitial macrophage accumulation in lupus nephritis and diabetic nephropathy.48,49 Interferon-inducible protein 10 (IP-10), also known as CXC-chemokine ligand 10 (CXCL10), is produced by many renal cell types and is a soluble chemoattractant for activated T cells. Urine IP-10 levels are increased in patients with diabetic nephropathy and renal allograft rejection.50,51 In addition, urine levels of IP-10 correlate with the incidence of renal allograft rejection and predict allograft function.52 CXC-chemokine ligand 16 (CXCL16) is another chemoattractant for activated T cells, which correlates with T-cell accumulation in acute and chronic renal diseases. Transmembrane CXCL16 is present on glomerular and tubular cells during injury and is released after proteolytic cleavage by metalloproteinases. Urine levels of soluble CXCL16 are increased in patients with lupus nephritis or renal allograft rejection.53,54 Macrophage migration inhibitory factor (MIF) is a molecule that is produced at sites of inflammation and inhibits further macrophage migration in response to chemokines, thereby allowing macrophages to accumulate at the inflammatory site. MIF can also enhance the activity of macrophages and T cells at sites of injury. Increasing levels of MIF in urine correlate with kidney leukocyte accumulation and the severity of renal damage in proliferative forms of glomerulonephritis.55 In addition, elevated MIF levels in urine can predict episodes of acute renal allograft rejection and discriminate from cyclosporine nephrotoxicity.56

There are also other pro-inflammatory mediators that can indentify inflammation in the injured kidney. Vascular cell adhesion molecules-1 (VCAM-1) is expressed by renal vessels and some kidney cells during renal inflammation and facilitates transendothelial leukocyte migration. Some of this VCAM-1 is enzymatically cleaved and excreted into the urine. Urine levels of soluble VCAM-1 are elevated during active periods of anti-nuclear cytoplasmic antibody vasculitis and lupus nephritis,53,57 and are useful for determining the severity and type of renal allograft rejection.58 Interleukin-18 (IL-18) is a pro-inflammatory cytokine that is produced by leukocytes, vessels and kidney tubules. During acute renal injury, there is a substantial increase in IL-18 production by tubules. Elevated urine levels of IL-18 are a relatively sensitive and specific marker of acute tubular necrosis (ATN) and delayed graft function in the post ischaemic kidney.59 Urine levels also correlate with disease activity in idiopathic nephritic syndrome.60 Tumour necrosis factor receptor-1 (TNFR1) is one of the major receptors for the pro-inflammatory cytokine TNF-α, which is expressed on infiltrating leukocytes and some resident kidney cells during renal inflammation. The soluble form of TNFR1 is more stable and easier to detect in serum and urine than TNF-α and it can serve as a surrogate marker of TNF-α activity in kidney disease. Serum and urine levels of soluble TNFR1 are increased during acute and chronic renal inflammation and correlate with the progression of acute renal failure, lupus nephritis and diabetic nephropathy.50,53,61 Another recent inclusion to this family of biomarkers is soluble human leukocyte antigen-DR. Urine levels of soluble human leukocyte antigen-DR are a sensitive and highly specific marker of acute renal allograft rejection, which can be detected up to 5 days before the clinical signs of acute cellular or vascular rejection are evident.62


The development of renal fibrosis is dependent on excessive production of profibrotic growth factors and extracellular matrix, which can be detected in urine by ELISA. Transforming growth factor-β1 (TGF-β1) and connective tissue growth factor are two of the major growth factors that promote renal fibrosis. Urine levels of TGF-β1 and connective tissue growth factor increase with the progression of CKD;63–65 however, TGF-β1 is mostly excreted as an inactive complex, which requires brief acidification to permit activation and detection. Some profibrotic molecules that are induced by TGF-β1, such as TGF-β-inducible gene H3 (βig-H3) and plasminogen activator inhibitor-1, are also detectable in urine and can act as surrogate markers of renal TGF-β1 activity. Urine levels of βig-H3 are about approximately 1000 times greater than TGF-β1 in diabetic patients and can be detected before the onset of albuminuria,66 indicating that βig-H3 is an early and sensitive marker of renal fibrosis during diabetes. Urine excretion of plasminogen activator inhibitor-1 has been shown to correlate with renal injury and fibrosis in patients with diabetic nephropathy and progressive chronic glomerulonephritis.67,68

Collagen type IV is a major component of kidney extracellular matrix, which is increased during the progression of renal fibrosis. Urine excretion of collagen IV is elevated in patients with IgA nephropathy and diabetic nephropathy and correlates with declining renal function.69,70 In addition, urine levels of collagen IV correlate with glomerular matrix accumulation and declining renal function in animal models of kidney disease.71 In contrast, serum levels of collagen IV are not associated with the development of renal injury or loss of kidney function.72


Although reliable ELISA exists for most of the recently described renal biomarkers in serum and urine, this technique is limited to measuring a single marker per assay, which makes assessment of multiple biomarkers time-consuming and expensive. Recently, multiplex assay systems have been developed by Luminex ( and BD Biosciences (, which uses the principles of both ELISA and flow cytometry to simultaneously quantitate multiple antigens in biological fluids. In the Luminex assays, microspheres with unique spectral signatures are coupled with primary antibodies. The antigens binding to these microspheres are then labelled with biotinylated secondary antibodies and streptavidin coupled to another fluorochrome (phycoerythrin). The microspheres and antigens labelled with phycoerythrin are excited with lasers at different wavelengths and the emission signals are used to identify the antigen and the amount of antigen bound to the microsphere. This technique is theoretically capable of assessing up to 100 different antigens and requires small volumes of biological fluid (30 µL). The Luminex assay system has been used to assess multiple biomarkers in the urine of patients with renal allograft rejection and lupus nephritis.51,73 The advantages and technical considerations for multiplex assays have been recently reviewed by Leng et al.74


Recent advances in proteomic analysis incorporating mass spectrometry have led to the identification of novel biomarkers of kidney injury in urine, which include: (i) intact or fragmented proteins that are selectively increased or decreased in kidney disease; (ii) protein patterns that are indicators of specific types of kidney disease; and (iii) protein patterns that can predict the progression of acute or CKD. Proteomic studies of patient urine have identified exosomal fetuin-A as an early biomarker of acute kidney injury,75 cleaved forms of β2-microglobulin as markers of acute renal allograft rejection,76 and a ubiquitin fusion protein (UbA52) as a potential specific marker of diabetic nephropathy.77 Interestingly, one of these studies also found that a fragment of degraded ubiquitin was specifically absent in urine from patients with diabetic nephropathy.77 Other researchers have focussed on urine proteomic patterns as a means to predict the progression of kidney diseases with high sensitivity and high specificity. A urinary polypeptide pattern has been shown to distinguish IgA nephropathy from normal controls (90% specificity) and from patients with membranous nephropathy, minimal change disease, FSGS or diabetic nephropathy (100% specificity).78 Another urine proteomic study found that two proteins in a mass spectrometer signature can distinguish active and inactive lupus nephritis with 92% specificity.79 In addition, a clinical analysis has identified a 12 peak proteomic mass spectrometer signature that can predict cases of diabetic nephropathy in 74% of type 2 diabetic patients before the onset of microalbuminuria.80 Similarly, a more complex panel of 65 biomarkers has been shown to predict the development of diabetic nephropathy in patients with microalbuminuria (97% sensitivity) and differentiate from other chronic renal diseases (91% specificity).81 In this latter study, many of the urine biomarkers identified were fragments of collagen type I that were reduced in diabetic patients. One general concern with urine proteomic studies is that they can identify proteins as potential biomarkers when they have no known relationship to kidney injury, and this lack of connection to disease pathophysiology is a significant limitation.82


Recent advancements in molecular analysis have resulted in the identification of a wide range of potential serum and urine biomarkers for assessing renal function and injury and predicting the development of kidney disease. Many of these biomarkers can be grouped according to their association with a particular type of injury (e.g. podocyte or tubular injury) or a mechanism of damage (e.g. oxidative stress, inflammation, fibrosis). Understanding the relationships between these different biomarker categories may help us to better understand disease processes. In addition, future assay developments may result in the creation of multiplex assays that target panels of biomarkers according to these specific categories. Such assays may prove useful for determining the nature of the renal injury or the stage of disease in patients. They may also help to better determine the most appropriate intervention therapies for patients and the efficacy of novel or established therapies for targeting specific disease processes. Biomarker panels could also be used as surrogate end points in clinical trials, which might speed up the clinical evaluation of new drugs.

Most of the serum and urine biomarkers described in this review are not unique to humans and can be detected in rodent models of kidney disease using similar assay systems. The ability to reliably measure these biomarkers in serum and urine samples is critically dependent on appropriate sample processing, which can significantly affect findings. Strict protocols need to be established for sampling and sample handling to minimize the variations in biomarker detection that are due to these procedures. After collection, serum and urine samples should be analysed immediately or frozen in aliquots. If urine samples are being collected over a timed period (e.g. a 24 h collection), protease inhibitors may need to be added to avoid degradation of protease sensitive molecules. In addition, frozen samples should be analysed at the first thawing, as repeated freeze-thaw cycles can result in the loss of some protein biomarkers by cryoprecipitation.

There is mounting evidence from small clinical studies that the progression of kidney diseases may be predicted by evaluating a combination of serum and urine biomarkers together with other risk factors such as age and hypertension. In the future, this analysis process may also include urine proteomic patterns and genetic biomarkers. However, larger clinical studies will be required to compare panels of biomarkers and achieve agreement on which combination offers the most useful and cost-effective clinical information.


GH Tesch is supported by a Career Development Award from the National Health and Medical Research Council of Australia, Kidney Health Australia and the Australian and New Zealand Society of Nephrology.