Understanding Kidney Function Assessment: The Basics And Advances

for more information:Ali.ma@wecistanche.com

Houry V. Puzantian, et al

Abstract Purpose:

Multiple kidney function assessment modalities are available, but their appropriateness is constantly questioned. This review provides practitioners with an in-depth understanding of kidney function assessment methods, their clinical utility, and comparisons.

Data sources: PUBMED search was conducted by relevant subject headings.

Conclusions: Glomerular filtration rate (GFR) is the best indicator of kidney function. Exogenous compounds like inulin help measure GFR, but endogenous substances (like creatinine) are more convenient, although exhibiting greater variability. Cystatin C is advocated as a functional marker; its clinical significance is under study. Proteinuria adds value to GFR estimation. There are commonly used equations estimating GFR like the creatinine-based Cockcroft–Gault and the modification of diet in renal disease. The new creatinine-based Chronic Kidney Disease Epidemiology Collaboration (CKDEPI) equation demonstrates higher accuracy of patient classification in earlier stages of the disease. Recently, the Chronic Renal Insufficiency Cohort (CRIC) study has devised an equation combining serum creatinine and cystatin C in longitudinal modeling of kidney function.

Implications for practice: Current GFR estimation methods have limitations, and are useful for populations they have been tested in. Practitioners should be well informed on emerging equations that provide greater accuracy in CKD diagnosis; this would help implement appropriate prevention and intervention strategies.

Keywords: Kidney function tests; kidney diseases; glomerular filtration rate; creatinine

cistanche for kidney function

Click to cistanche tubulosa benefits and side effects and Cistanche for kidney function

Kidney function assessment is an integral component of diagnosing and initiating prevention and treatment of chronic kidney disease (CKD). It is estimated that approximately 26 million people in the United States have CKD (Coresh et al., 2007). Apart from advancement into kidney replacement therapy, these patients are at high risk for cardiovascular events and mortality (Collins et al., 2003) necessitating close medical follow-up. Determination of kidney function also aids in clinical decisions related to fluid administration and drug dosing; it is pivotal in the prevention of adverse effects resulting from diagnostic and therapeutic procedures such as those requiring the use of intravenous contrast medium. Hence, appropriate assessment of kidney function plays an essential role in inpatient care. Over the years, the development of multiple approaches to kidney function assessment has raised concerns regarding their limitations and clinical applicability. The purpose of this review is to provide practitioners with an in-depth understanding of kidney function assessment methods, their appropriate clinical utility, and how they compare.

Methods of measurement of kidney function

Glomerular filtration rate (GFR) is accepted as the best overall indicator of kidney function. Reduced GFR could indicate either primary kidney disease or a secondary problem such as decreased kidney perfusion or drug toxicity (Stevens & Levey, 2005). It is, in fact, a persistently low GFR that provides a good index of declining kidney function (National Kidney Foundation [NKF], 2002), and it would be important to note whether GFR is changing or whether it is stable. Currently, one way of defining CKD is still by the level of GFR (Table 1; NKF, 2002). The severity of CKD is also determined by the level of GFR (Table 2).

GFR is defined as the amount of plasma filtered through the glomeruli per unit of time and represents the sum filtration rate of all functional nephrons. Normal GFR varies according to age, gender, race, and body size (Stevens & Levey, 2005). GFR is approximately 120– 130 mL/min/1.73 m2; it declines with age, is lower in females, is higher in African Americans, and varies with skeletal muscle diseases and amputations. GFR can either be measured or estimated by equations.

GFR cannot be measured directly. It is measured through the urinary clearance of a filtration marker. An ideal marker would be a freely filtered inert substance, not metabolized, secreted, or reabsorbed through the kidney (Traynor, Mactier, Geddes, & Fox, 2006). Exogenous and endogenous markers have been used.

Exogenous markers

Inulin is the gold standard filtration marker but is expensive and cumbersome to measure (Stevens & Levey, 2005). To achieve steady plasma levels, inulin bolus and infusion are required; Some blood and urine samples are needed to estimate inulin clearance.

Investigators have also used radioisotopic compounds: iodine-125-iothalamate, chromium-51-ethylenediaminetetraacetic acid, and technetium-99m-diethylenetriamine pentaacetic acid. These require handling precautions, are expensive, overestimate GFR, exhibit prolonged elimination in advanced kidney disease, and are inappropriate for use in pregnancy (Rahn, Heidenreich, & Bruckner, 1999; Traynor et al., 2006) or in patients with difficulty emptying the bladder.

Nowadays, radiocontrast agents (nonradioactive) like iohexol, iothalamate, and diatrizoate meglumine are available and are considered safer than radioactive agents. Iohexol is advocated as a marker with a clearance comparable to that of inulin. It can be measured in plasma, serum, and urine using high-performance liquid chromatography. While these methods have potential, the bolus administration of substances and required plasma tracings render them undesirable.

cistanche for kidney function

Endogenous markers

Difficulties encountered with the measurement of exogenous markers have led to the relentless evaluation of filtration processes of endogenous substances.

Urea. Urea, an end product of protein catabolism, is synthesized in the liver mainly from dietary protein intake. Production increases with high protein diets, medications like corticosteroids and tetracyclines, or conditions like trauma, gastrointestinal hemorrhage, infections, heart failure, and acute kidney failure (Stevens & Levey, 2005; Traynor et al., 2006). Although urea is freely filtered at the glomerulus, 40%–50% is reabsorbed at proximal and distal tubules, underestimating GFR. Volume depletion and antidiuresis increase urea reabsorption, with a greater decrease in urea clearance than in GFR. Extracellular volume expansion and diuresis increase urea clearance are more than the GFR. These factors create intraindividual and interindividual differences in urea generation and excretion, rendering urea an unreliable marker of kidney function.

Creatinine clearance.

CrCl is based on urinary creatinine, urinary volume in a 24-hour period, and serum creatinine (SCr) levels, indicating creatinine excretion per day: CrCl (ml/min) = [urine creatinine (mg/mL) × 24-hour volume (mL)]/[creatinine in blood (mg/mL) × 24 × 60 min]. In addition to the glomerular filtration of creatinine, renal tubules secrete creatinine; hence measurements of CrCl can overestimate GFR. For example, people with high body mass index (BMI) exhibit a BMI-associated increase in the tubular secretion of creatinine, with CrCl overestimating true GFR (Sinkeler et al., 2011). In earlier kidney disease, this overestimation is a systematic error shifting results in the same direction, and CrCl continues to be useful in monitoring kidney function changes in the same patient. However, as GFR decreases with advanced disease, a variable increase is observed in the proportion of creatinine secretion to filtration (Stevens & Levey, 2005; Traynor et al., 2006). Therefore, CrCl is an inaccurate indicator of kidney function at lower GFR levels, underestimating the severity of kidney disease. It is suggested that CrCl be measured with cimetidine, which inhibits the tubular secretion of creatinine; leading to better GFR estimation (Walser, 1998). Daily variations in creatinine excretion should also be considered. Furthermore, 24-h urine collection is cumbersome, and error-prone because of specimen losses and overcollections from failing to flush the first voided sample. However, CrCl continues to be recommended even for conditions impacting creatinine concentration where its daily production is difficult to estimate: for example, vegetarian diet, malnutrition, obesity, skeletal muscle diseases, paraplegia, quadriplegia, or amputation, and pregnancy (Fawaz & Badr, 2006). Because CrCl tends to overestimate true GFR, and urea clearance underestimates GFR, some have recommended averaging values of both measured at the same time to obtain a closer estimate of kidney function for patients anticipated to be in stage 4 or 5 CKD (Almond, Siddiqui, Robertson, Norrie, & Isles, 2008).

Cistanche can improve kidney function

Serum creatinine.

SCr has been the most thoroughly investigated glomerular filtration marker. Creatinine is produced by skeletal muscle breakdown and is contained in cooked meat. Several factors influence SCr. Advancing age, female gender, and white race are associated with lower SCr because of lower muscle mass than in younger age, male gender, and black race (Jones et al., 1998). Changes in body habits such as amputations, and variations in dietary intake such as vegetarian diet or creatinine supplements cause alterations in SCr. Creatinine is freely filtered at the glomerulus and not reabsorbed, but 10%–15% is secreted at the tubules and extrarenal elimination occurs through the gastrointestinal tract (Fawaz & Badr, 2006; Stevens & Levey, 2005; Traynor et al., 2006).

Kidney parenchymal disease may initially lead to hypertrophic and hyperinflation compensatory mechanisms in functional nephrons preventing SCr increases, thus masking kidney function deterioration (Shemesh, Golbetz, Kriss, & Myers, 1985). A substantial decrease in GFR from 120 to 80 mL/min/1.73 m2 is accompanied by only a small rise of SCr from 0.8 to 1.2 mg/dL (Figure 1; Inker & Perrone, 2012). Therefore, SCr is an insensitive marker of kidney function in early kidney disease.

As kidney disease progresses, creatinine is hypersecreted from functional nephron tubules leading to an overestimated GFR. Prior to creatinine standardization (Myers et al., 2006), there were inaccuracies in creatinine measurement and variations in assay calibration among and within laboratories (Coresh et al., 2002; Murthy, Stevens, Stark, & Levey, 2005) causing inappropriate kidney function evaluation. For all the above reasons, other kidney function markers have been sought.

Cystatin C.

Serum cystatin C, a relatively novel marker, is potentially superior to SCr under certain conditions. Cystatin C is a protein belonging to the cysteine protease inhibitor superfamily, is produced by nucleated cells, filtered at the glomerulus, reabsorbed, and metabolized at tubules (Madero, Sarnak, & Stevens, 2006). Its reabsorption, metabolism, and extrarenal excretion hinder proper measurement of its urinary excretion.

Serum cystatin C exhibits greater intraindividual variability than SCr (Madero et al., 2006). Factors associated with cystatin C are height, weight, smoking, diabetes, white blood cell count, thyroid function status, corticosteroids, and inflammation such as increased high sensitivity C-reactive protein levels (Knight et al., 2004; Stevens et al., 2009). Unlike SCr, cystatin C is not affected by muscle mass and dietary factors; it is a gene product and is generated continuously (Abrahamson et al., 1990). Although cystatin C was thought to be independent of age and gender (Laterza, Price, & Scott, 2002) recent reports indicate older age and male gender may be associated with higher cystatin C levels (Knight et al., 2004).

Cystatin C is a more sensitive marker than creatinine in detecting an early reduction in kidney function (Coll et al., 2000); however, at lower levels of GFR (≤70 mL/min/1.73 m2), SCr-based assessments may perform better. In the presence of promising evidence on cystatin C, a review of multiple studies on GFR estimation revealed serum cystatin C to be either equivalent or superior to SCr (Dharnidharka, Kwon, & Stevens, 2002); conclusive evidence is still pending (Prigent, 2008).

Prediction equations for estimation of kidney function

GFR-estimating equations provide robust GFR estimates by the inclusion of demographic and physiology variables affecting endogenous substances like Cr. Critiques emphasize the careful interpretation of results and classification schemes derived from these assessment tools (Glassock & Winearls, 2008). The performance of equations is evaluated through measures of bias, precision, and accuracy (Stevens, Zhang, & Schmid, 2008). Bias is the mean difference between measured GFR (mGFR) and estimated GFR (eGFR). Precision refers to variation or spreads around that mean difference. Accuracy attests to both bias and precision. Estimates with high accuracy have low bias and high precision (Figure 2). Accuracy is often judged by the value of P30 (accuracy within 30%), which is the percentage of eGFRs within 30% of mGFR.

Figure 2 Bias, precision, and accuracy

Cockcroft–Gault equation

The Cockcroft–Gault equation (Table 3) is based on a study performed on 249 hospitalized male patients (corrected for females), ages 18–92 years (Cockcroft & Gault, 1976). It includes actual body weight, which should theoretically take differences in muscle mass into account. The goal was to estimate CrCl without a 24-h urine collection.

The equation is based on SCr; computed CrCl values vary from mGFR because of SCr measurement errors. Recalibration to the original assay cannot be performed because the laboratory methods used to derive the formula have been abandoned (Stevens & Levey, 2005). The equation can overestimate CrCl in obesity and fluid overload states, where the “actual” weight may not clearly predict muscle mass (Traynor et al., 2006). However, despite its limitations, the equation is useful for tracking changes in kidney function and for drug dosing (FDA labeling requirements).

Modification of diet in renal disease (MDRD) study equation

The MDRD equation (Table 3), recommended by the NKF (2002), is in wide clinical use. It resulted from the MDRD study investigating 1628 subjects with advanced nondiabetic kidney disease (Levey et al., 1999). The original equation included age, gender, race, SCr, serum urea nitrogen, and albumin concentrations. The investigators reported that eGFR by the MDRD equation did not systematically deviate from concurrent mGFR, and was thus unbiased. In addition, 91% of eGFRs predicted by the equation were within 30% of concurrent mGFR values (Levey, Greene, Kusek, & Beck, 2000), so it is reasonably accurate.

The calibration to a standardized SCr based on gold standard methodology has been highly recommended for the proper use of GFR-estimating equations (Coresh et al., 2002; Myers et al., 2006). The incorporation of standardized SCr in the MDRD equation provides more accurate eGFRs than unstandardized SCr measures (Levey et al., 2006, 2007).

While the Cockcroft–Gault relies on weight, the MDRD equation is adjusted for body surface area accounting for variations in muscle mass with certain diseases or amputations. The MDRD equation outperforms the Cockcroft– Gault formula in older age, obese patients (Fares et al., 2004), and diabetics (Poggio, Wang, Greene, Van Lente, & Hall, 2005). The MDRD equation, like the Cockcroft– Gault, is less accurate in early kidney disease; it is biased toward underestimating kidney function (Poggio, Wang, et al., 2005). Its use in hospitalized ill patients needs further verification (Poggio, Nef, et al., 2005). The MDRD has not been tested in children, pregnancy, or extremes of body size.

Cystatin C-based equations

In recent years, cystatin C-based GFR-estimating equations (Table 3) have been developed (Madero et al., 2006; Stevens et al., 2008). Although there has been a lack of homogeneity in eGFR for similar cystatin C levels by using different equations, some studies report improved GFR estimation with cystatin C-based equations (Tanaka, Suemaru, & Araki, 2007). The latter reflects kidney function more accurately than creatinine-based equations in populations producing low levels of creatinine like elderly, children, kidney transplant recipients, and cirrhosis patients. Despite these attempts, comparative studies on cystatin C and creatinine-based equations remain inconclusive (Stevens, Padala, & Levey, 2010). Using both cystatin C and SCr, however, has been shown to provide better estimates of GFR than equations utilizing the markers separately; the percentage of eGFRs falling within 30% of mGFRs increased from 80.4% in creatinine-based equations to 89% with a creatininecystatin C combination equation (Stevens et al., 2008; Tidman, Sjostrom, & Jones, 2008).

Recommendations regarding the clinical use of cystatin C-based equations are still pending (Prigent, 2008). It is recommended that the calibration of cystatin-C assays be standardized, and cystatin-C-based equations be further validated in different populations.

CKD epidemiology collaboration (CKD-EPI) equation

The CKD-EPI equation (Table 3) was recently developed from a compiled dataset including people with and without kidney disease. The main objective was to achieve greater accuracy at higher GFRs as compared to the MDRD equation (Levey, Stevens, et al., 2009). The population included mostly whites and African Americans. Asians had not been adjusted for in the equation, a correction coefficient (0.813) was calculated for use of CKD-EPI in a Japanese cohort (Horio, Imai, Yasuda, Watanabe, & Matsuo, 2010).

The new CKD-EPI equation outperforms the recommended MDRD. In the original study, lower bias, higher accuracy, and precision were obtained with the CKDEPI equation than with the MDRD (p < .001) (Levey, Stevens, et al., 2009), mostly in patients with eGFR ≥ 60 mL/min/1.73 m2. With CKD-EPI, a significantly greater percentage (p < .001) of eGFRs were within 30% of mGFR than that of the MDRD; however, the authors still viewed the accuracy as suboptimal. Moreover, using CKD-EPI more patients would be classified as stage 2 who would otherwise be classified as more advanced stage 3 cases in a false-positive manner, by the use of the MDRD. This indicates that CKD-EPI has a lower bias compared to the MDRD equation. CKD-EPI equation suggested a CKD prevalence of 11.5%; lower than that (13.1%) obtained by the MDRD equation (Levey, Stevens, et al., 2011). Decreased prevalence rates have also been obtained with the CKD-EPI equation in other studies in the United States, Australia, and Japan (Horio et al., 2010; Matsushita, Selvin, Bash, Astor, & Coresh, 2010; White, Polkinghorne, Atkins, & Chadban, 2010). In a Singaporean study of Chinese, Malays, and Indians, similar prevalences were obtained by the two equations (Sabanayagam, Wong, & Tai, 2009). This discrepancy could be because of the influence of other factors like differences in creatinine assays, sample characteristics, muscle mass, and diet (Levey, Stevens, et al., 2011).

Recently, CKD-EPI investigators reported on a newly developed combined creatinine-cystatin C equation that is more accurate in CKD classification than equations using either marker alone (Inker et al., 2012).

Cistanche can improve kidney function

Chronic Renal Insufficiency Cohort (CRIC) study equation

In an effort to improve the estimation of GFR, the CRIC study (Feldman et al., 2003) reports a new GFRestimating equation to include both SCr and cystatin C, in addition to age, gender, and race (Anderson et al., 2012). Using this equation, 89% of eGFRs fall within 30% of mGFRs. This equation was developed internally for kidney function assessment in CRIC and will be utilized to monitor the progression of CKD in that study cohort. External validation would be required to determine its clinical utility and generalizability to different populations.

Considering all equations for kidney function estimation, future studies should pursue new markers to improve the accuracy of GFR estimates and explore methods to evaluate change in GFR over time.

Proteinuria as an indicator of kidney disease

Healthy kidneys excrete small amounts of protein. Persistently increasing levels of the urinary protein indicate kidney damage as per NKF Kidney Disease Outcomes Quality Initiative (NKF KDOQI) practice guidelines (NKF, 2002). The NKF CKD staging system views proteinuria and eGFR as separate markers of kidney disease. A recent comparison, considering proteinuria in combination with eGFR, was more likely than the NKF staging system to correctly classify individuals for kidney disease outcome: doubling of creatinine by end of follow-up, dialysis initiation, or kidney transplant (Tonelli et al., 2011).

The urinary albumin concentration is, in turn, an independent predictor of all-cause mortality in the general population (Hillege et al., 2002; Matsushita, van der Velde et al., 2010). Furthermore, increased albuminuria is associated with both cardiovascular disease and mortality in patients with a history of hypertension, diabetes, or cardiovascular disease (van der Velde et al., 2011); microalbuminuria seems to reflect diffuse endothelial injury (Glassock, 2010). Albuminuria is a significant prognostic marker and is advocated by the current expert consensus for integration in the eGFR-based kidney disease staging process (Levey & Coresh, 2012). Albumin to creatinine ratio (ACR) > 17 mg/g for men and >25 mg/g for women is considered high or very high (Levey, Cattran, et al., 2009), commensurate with CKD. Ongoing work involves the modification of global clinical practice guidelines by a Kidney Disease: Improving Global Outcomes (KDIGO) workgroup.

Conclusion

The appropriateness of kidney function assessment strategies is of concern. GFR is determined by measuring the excretion of exogenously administered or endogenous compounds, or by estimation equations. There are difficulties in measuring GFR by exogenous substances; endogenous markers have been the mainstay in kidney function assessment. Cohort studies like the CRIC have shown that eGFR performs as well as mGFR for common clinical endpoints related to kidney failure: anemia, acidosis, and increases in potassium or phosphate (Hsu et al., 2011). Urea is a marker that correlates best with advanced kidney disease stages, but it is not deemed a reliable marker because of high intraindividual and interindividual variability. CrCl could be utilized in patients with early kidney disease however, the 24-h urine collection period renders it error-prone and cumbersome for patients, and it is not a good indicator of GFR in advanced disease. SCr, once a promising filtration marker, is a relatively poor indicator of kidney function on its own. Several factors affect its generation, creatinine exhibits kidney tubular secretion, incoherence with GFR in early kidney disease, and potential for laboratory measurement errors. However, SCr is an essential component of GFRestimating equations. Cystatin C, a newer kidney function marker, seems to be equivalent to SCr in estimating GFR; its superiority to SCr is still debatable. Proteinuria is advocated as an important marker in combination with eGFR, potentially enabling practitioners to better evaluate kidney disease progression.

Multiple equations have been developed for GFR estimation using demographic and clinical variables. The MDRD equation, currently recommended by the NKF, has been shown to outperform the Cockcroft–Gault formula. The MDRD is a relatively robust equation that relies on standardized SCr in addition to demographic variables. Cystatin C has been integrated into GFR-estimating equations because of its high potential for proper kidney function assessment; however, comparative results with SCr-based equations are still inconclusive. Recently, the CKD-EPI equation has been recommended to replace MDRD in routine clinical practice (Levey, Stevens, et al., 2011; Stevens et al., 2010). Its higher accuracy in CKD classification of early disease helps redirect resources to ill patients and facilitates medical processes for low-risk patients. Furthermore, the CRIC study, based on a cohort of participants, has devised a new GFR-estimating equation to include both SCr and cystatin C (Anderson et al., 2012). Its clinical utility is yet to be explored.

Finally, the performance of GFR estimation methods is critical. Available equations exhibit limitations; their utilization would be most appropriate for populations in which they have been tested. Practitioners must consider non-GFR estimates, type of kidney disease, proteinuria, and urinary sediments inpatient evaluation. Further clarification is needed on how practitioners should apply GFR estimates obtained by different methods in various clinical settings. Moreover, practitioners should be vigilant in detecting upcoming equations that could assist in diagnosing CKD with greater precision, discerning the different stages, and intervening accordingly.

Cistanche can improve kidney function

Acknowledgments

The authors thank Dr. Barbara Riegel, Dr. Pamela Cacchione, and Ms. Justine Sefcik for their review and feedback on the manuscript

Understanding kidney function assessment

From: ' Understanding kidney function assessment: The basics and advances' by Houry V. Puzantian, et al

---Journal of the American Association of Nurse Practitioners 25 (2013) 334–341 C 2013 The Author(s) C 2013 American Association of Nurse Practitioners