Summary

Glomerular filtration rate (GFR) is the best index for assessing the overall renal function but it can only be determined indirectly. Measurement of serum creatinine is the simplest and the most widely used method for GFR assessment. Correct interpretation of serum creatinine levels requires knowledge about the age, sex, muscle mass and physiological states of individual patients and factors that may interfere with laboratory measurements. Determination of creatinine clearance (CrCl) using 24-hour urine sample is also used for GFR assessment. CrCl can overestimate GFR because of the renal tubular secretion of creatinine. Under or over collection of 24-hour urine can cause significant errors in CrCl calculations. The Cockcroft & Gault formula and the abbreviated MDRD equation can provide additional information to serum creatinine level alone. Understanding their limitations is required for accurate application of these equations in clinical practice.

Numerous disease processes and pharmaceutical agents have primary and secondary effects on the kidneys. A stepwise investigative approach is required to delineate the underlying cause of renal impairment. Physical examination, routine laboratory tests and some basic radiological examinations help divide the cause of renal impairment into prerenal, renal and postrenal forms. This information is very useful in guiding appropriate management and further investigations.

摘要

腎小球率過濾(GFR)是評估整體腎功能的最好指標，但只可能間接地測定，而量度血肌酐則是評估 GFR 的最簡單和最常用的方法。正確分析血清肌酐水平，需要了解病人的年紀、性別、肌肉多少、生理狀況和可以影響實驗結果的原因。24小時尿肌酐清除率(CrC1)也可以用來評估GFR，高估GFR常因為腎小管分泌肌酐，採集24小時小便樣本不準確(過少或過多)，都可導致計算CrC1時出現顯著的誤差。Cockcroft & Gault 的公式和簡化MDRD方程式可對血肌酐分析提供額外資料。在臨床上使用時，應了解其局限性。

眾多疾病和藥物對腎臟都有直接和間接的影響。需要採用階梯式的調查方法去描述腎功能損害的原因。通過身體檢查、例行實驗室化驗以及基本放射學的檢查可以將病因分成腎前性的、腎性和腎後性的三種。這對於正確的治療和進一步檢查非常重要。

---

Introduction

The kidney performs numerous functions - elimination of waste products, water homeostasis, regulation of electrolytes and acid-base balance, synthesis of 1,25-dihydroxy-vitamin D, renin and erythropoietin. Assessment of these functions of the kidney requires appropriate tests, some of which are performed in a chemical pathology laboratory.

"Renal Function Test" or RFT is frequently requested as a routine screening test, either for diagnosis and monitoring of renal diseases, drug dosage adjustments or as part of the basic workup of patients. The definitions of RFT and the tests included thereof may be different for different laboratories. RFT is a test profile that usually includes urea, creatinine, sodium, potassium and occasionally chloride and bicarbonate. Sodium and potassium are the two most abundant cations in our body. Their changes can be the results of numerous diseases, while a renal disease is just one of these. Similarly, the levels of chloride and bicarbonate are disturbed in acid-base disorders, which can originate from the kidneys or other organ systems in the body. Therefore, any changes in serum (or plasma) levels of sodium, potassium, chloride or bicarbonate, as reflected in the results of a RFT, are not necessarily due to problems of the kidneys. In this article, we will concentrate on the interpretation of urea and creatinine and other laboratory tests used in the assessment of glomerular filtration rate (GFR). Common patterns of RFT will be illustrated by clinical cases.

Urea and creatinine

Urea and creatinine are the most commonly used serum markers to screen for renal disease. Both analytes are metabolic end products and are excreted almost exclusively by the kidneys. Factors that affect the measured serum levels of urea and creatinine are listed in Table 1.

Urea is derived from the breakdown of dietary and endogenous proteins. It is freely filtered at the glomerulus and undergoes 50% reabsorption at the renal tubules. The amount reabsorbed is increased during dehydration and in conditions with low urine flow rate. Low urea levels are seen in people on a low protein diet.

Creatinine is mainly derived from the metabolism of creatine in skeletal muscles. Its serum level is proportional to the total muscle mass. Lower levels are seen in infants and children, the elderly, and in female than in male. Therefore, we need to consider the age, sex and muscle mass of an individual for correct interpretation of serum creatinine results. The following example can illustrate this point:

Case 1:

What is the significance of a serum creatinine of 120 mmol/L in:-

(a) 70 kg weight lifter?

(b) 40 kg elderly woman?

It is very common for laboratories to provide a single reference interval for serum creatinine (usually 60 - 120 mmol/L) for all subjects. Underestimation of renal impairment may result if this reference interval is applied to all individuals without consideration of other physiological factors that can affect the serum creatinine level. In the above example, a serum creatinine of 120 mmol/L reflects normal renal function in Case 1(a) but mild to moderate renal impairment in Case 1(b).

Most laboratories measure creatinine by methods based on the Jaff reaction. In this reaction, creatinine reacts with alkaline picrate reagent to form an orange compound. The changes in absorbance at specific wavelengths reflect creatinine concentration. Despite its common use, the Jaff reaction is subjected to a number of interferences.^1,2 Ketone bodies and some cephalosporins can lead to positive interference whereas bilirubin may lead to negative interference.

Case 2:

A 39-year old woman was given intravenous cefoxitin. The following results were noticed.

Serum		Admission		Day 1
								Reference interval
Na		137		135		mmol/L			135	-	149
K		3.8		3.9		mmol/L			3.4	-	5.2
Urea		4.3		4.5		mmol/L			3.3	-	7.0
Creatinine 97 559									60	-	120

There was a marked increase in serum creatinine concentration. In view of the possibility of acute renal failure, an urgent ultrasound scan of the kidneys was performed and showed no evidence of obstruction. In contrast to creatinine, urea did not show any significant change. A positive interference in the creatinine measurement by cefoxitin was suspected.³ Creatinine measurement was then repeated by an enzymatic method, which showed a level of 107 .

This kind of interference is more commonly seen in hospitalised patients who have been given cephalosporins by intravenous route. Usual oral doses rarely result in high enough serum drug concentrations that can lead to significant false elevation of serum creatinine levels. Enzymatic assays for creatinine measurement are associated with less interference but are more costly than those assays based on the Jaff reaction.

Glomerular filtration rate

Estimates of GFR are the best indices of the level of renal function. Chronic renal failure is preceded by a progressive decline in GFR in most forms of renal diseases. Estimation of GFR also helps to determine the proper dosages of renal excreted drugs. Conventionally, GFR is corrected for body surface area, which in the average adult is approximately 1.73m.² The corrected GFR for healthy adults is 80 - 120 ml/min/1.73m² and this value decreases with age at a variable rate.

GFR can be determined indirectly by measuring the renal clearance of a substance which (1) has stable concentration in blood, (2) is freely filtered at the glomerulus, and (3) is not reabsorbed, secreted, synthesised or metabolised by the renal tubules. The clearances of exogenous substances like inulin, ¹²⁵I-iothalamate and ⁵¹Cr-EDTA are considered the best markers of GFR. These tests, however, are too tedious and expensive for routine clinical use.

In routine clinical practice, the most commonly used marker of GFR is creatinine clearance (CrCl).⁴ Calculation of CrCl usually involves the collection of 24-hour urine and a random serum for creatinine measurements. The equation for CrCl calculation is shown in Figure 1. CrCl tends to overestimate GFR because creatinine is not only filtered at the glomerulus, but is also secreted by the proximal tubules. The day-to-day variation in creatinine excretion also contributes to the imprecision of CrCl determination. However, most of the inaccuracy of CrCl determination in 24-hour urine comes from improper urine collection. Both incomplete collection and over-collection are frequently encountered.

The limitations of using serum creatinine level as a marker of GFR have been discussed in the previous section. Another important point to note is that serum creatinine is not a sensitive marker for mild renal impairment. GFR has to drop to at least 50% of normal before we can see a rise in serum creatinine level above the upper reference limit. The curvilinear relation between serum creatinine and GFR is shown in Figure 2.

Glomerular filtration rate and CrCl can be estimated using equations based on serum creatinine level. It is generally agreed that these equations provide more accurate estimations of the degree of renal impairment than by serum creatinine level alone.⁵ The Cockcroft & Gault formula (Figure 1) is the most commonly used equation for CrCl estimation.

The abbreviated MDRD equation is gaining popularity in recent years. This equation is derived from the Modification of Diet in Renal Disease Study and it estimates GFR based on the serum creatinine level, age, gender and race (Figure 2).^6,7 This equation has been shown to have superior performance than the Cockcroft & Gault formula and measured CrCl in 24-hour urine in patients with chronic kidney disease. Around 90% of the estimated values were within 30% of the true GFR as determined by the gold standard methods and has a correlation coefficient (R²) of 0.89 against these reference methods.^6,7 However, the abbreviated MDRD equation tends to underestimate GFR in healthy persons and patients with mild degree of renal impairment.⁸ Another concern is whether this equation can be applied accurately to Chinese patients with chronic kidney disease.⁹ The latter awaits further validations.

The accuracy of these GFR-predicting equations diminishes (1) when renal function is changing rapidly, (2) at very high and very low GFR, (3) when normal body composition is disturbed (e.g. amputation), and (4) in individuals with variations in dietary intake (e.g. vegetarian diet, creatine supplements). The creatinine assays also affect the accuracy of these GFR-predicting equations. The commercial creatinine assays exist in the market differ in their calibrations. Therefore, different assays can give rise to different creatinine values for the same sample. For accurate applications of the GFR-predicting equations, the creatinine assay must be calibrated to the laboratory that developed the equation.

In recent years, some national and international clinical practice guidelines have been published which recommended clinicians to use estimating equations for GFR to assess kidney function.^5,10,11 However, it is crucial for the clinical users to understand the limitations of these equations, to interpret them in the proper clinical context and to apply them judiciously in their clinical practice.

Common patterns of abnormal RFT

Renal failure is often divided into acute and chronic types. Chronic renal failure describes the progressive decline in renal function over long periods of time. Acute renal failure, on the other hand, is defined as any sudden reduction in renal function. Acute renal failure can be further divided into prerenal, renal and postrenal forms. Prerenal form of acute renal failure results from reduced renal perfusion, which occurs in patients with acute circulatory failure. Postrenal renal failure is a result of urinary tract obstruction. Common causes include bladder stones and hypertrophy of the prostate gland. With early therapeutic intervention, prerenal and postrenal causes of renal impairment are usually fully reversible.

When a patient is found to have deranged renal function, further investigations are always required to delineate the underlying cause as numerous disease processes and pharmaceutical agents have primary or secondary effects on the kidneys. Some simple laboratory tests (e.g. calcium, urate, glucose, complete blood count, urine microscopy, etc) and radiological examinations (e.g. "KUB", ultrasound) can give valuable clues in most circumstances. More specialised tests should follow if necessary.

Case 3:

A 75-year old woman with type II diabetes mellitus, hypertension and recurrent episodes of minor stroke. RFT was checked at a routine follow-up.

Serum				Latest result				Reference interval
		Previous result checked 3 months before
Sodium		131		129		mmol/L			135	-	149
Potassium		4.7		5.1		mmol/L			3.4	-	5.2
Urea		29.7		26.4		mmol/L			3.3	-	7.0
Creatinine		373		451					60	-	120
Bicarbonate		21		19		mmol/L			22	-	26
Calcium		2.09		2.16		mmol/L			2.20	-	2.60
Phosphate		2.39		2.45		mmol/L			0.80	-	1.40
CrCl by Cockcroft & Gault formula (Weight = 42 kg)		7.47		6.17		ml/min					---

This patient had chronic renal failure. In general, changes in potassium, bicarbonate, phosphate and calcium levels do not become evident until the GFR falls to below 20 - 25 ml/min. Therefore, if the above electrolytes and acid-base disturbances were found in a patient with only mild degree of renal impairment, another disease process should be looked for.

Case 4:

A 79-year old woman presented with fever and drowsiness. She was an old-age home resident.

Serum						Reference interval
Sodium		160		mmol/L			135	-	149
Potassium		3.4		mmol/L			3.4	-	5.2
Urea		21.0		mmol/L			3.3	-	7.0
Creatinine		125					60	-	120

The severe hypernatraemia of this patient was due to dehydration, which was the result of poor oral intake and increased insensible water loss (fever). This patient also had renal impairment as reflected by the elevated serum urea and creatinine levels. The relatively greater increase in urea than creatinine supported the diagnosis of prerenal renal failure. Urea-to-creatinine ratio is a useful marker in this clinical situation. In normal individuals, the ratio is around 0.04 - 0.06 (mmol:mmol). Increased values suggest either an increase in urea production or an increase in tubular reabsorption of urea due to prolonged tubular transit from marked dehydration or reduced renal perfusion. The urea-to-creatinine ratio of this patient was 0.17.

Case 5:

A 33-year old engineer with acute gastrointestinal bleeding.

Serum						Reference interval
Sodium		139		mmol/L			135	-	149
Potassium		4.0		mmol/L			3.4	-	5.2
Urea		12.9		mmol/L			3.3	-	7.0
Creatinine		88					60	-	120

A mild to moderate elevation of serum urea is a common finding after an acute gastrointestinal bleed. Blood in the gastrointestinal tract acts like a large protein meal that increases urea production. Decreased renal blood flow due to hypovolaemia or blood loss can also contribute to this biochemical change.

Case 6:

A 50-year old woman presented with fever, nausea and skin rash for four days. Her blood pressure and hydration status was normal. Urinary tract obstruction has been ruled out by ultrasound scan of the urinary tract. Her past medical history was unremarkable. About one week ago, she had consumed multiple over-the-counter medications and proprietary herbal medicine for symptoms of upper respiratory tract infection. The following biochemistry results were obtained.

Serum						Reference interval
Sodium		137		mmol/L			135	-	149
Potassium		3.8		mmol/L			3.4	-	5.2
Urea		11.0		mmol/L			3.3	-	7.0
Creatinine		220					60	-	120

The clinical diagnosis was acute renal failure due to intrinsic renal disease. Renal biopsy confirmed acute interstitial nephritis, which was most likely drug-related. The renal function of this patient returned to normal spontaneously three weeks later.

Case 7:

A 65-year old man with benign prostate hyperplasia. His urine output decreased in recent two days. Examination of the abdomen found a palpable bladder.

Serum						Reference interval
Sodium		142		mmol/L			135	-	149
Potassium		5.2		mmol/L			3.4	-	5.2
Urea		25.6		mmol/L			3.3	-	7.0
Creatinine		510					60	-	120

This is a case of postrenal renal failure due to urinary tract obstruction by the hyperplasic prostate. Dramatic improvement of renal function and complete normalization may occur if the obstruction is relieved early enough.

Key messages

1. Result of a RFT profile is neither a precise index of renal disease nor a sensitive marker of renal function impairment.
2. Serum urea and creatinine concentrations are determined by the rates of production and elimination. Correct interpretation of serum creatinine levels requires knowledge about the age, sex, muscle mass and physiological states of individual patients and factors that may interfere with its laboratory measurements.
3. GFR is the best index for assessing the overall renal function.
4. Serum creatinine is not a sensitive marker of mild decrease in GFR.
5. Collection error and inconvenience to patients are the major drawbacks of using measured creatinine clearance based on a 24-hour urine specimen as an estimation of GFR.
6. The Cockcroft & Gault formula and the abbreviated MDRD equation can provide additional information to serum creatinine level alone. Understanding of their limitations is required for accurate application of these equations in clinical practice.

---

---

References