Clearance

An estimate of the GFR can be made by measuring the urinary excretion of a substance that is completely filtered from the blood by the glomeruli and is not secreted, reabsorbed or metabolized by the renal tubules. Experimentally, inulin (a plant polysaccharide) has been found to meet these requirements. The volume of blood from which inulin is cleared or completely removed in 1 min is known as the inulin clearance, and is equal to the GFR.

Measurement of inulin clearance requires the infusion of inulin into the blood and is not suitable for routine clinical use. The most widely used biochemical clearance test is based on measurements of creatinine in plasma and urine. This endogenous substance is derived mainly from the turnover of creatine in muscle and daily production is relatively constant, being a function of total muscle mass. A small amount of creatinine is derived from meat in the diet. Creatinine clearance is calculated using the formula:

   (4.1)

U = urinary creatinine concentration (µmol/L)

 = urine flow rate (mL/min or (L/24 h)/1.44)

P = plasma creatinine concentration (µmol/L)

Creatinine clearance in adults is normally of the order of 120 mL/min, corrected to a standard body surface area of 1.73 m². It should be noted that the clearance formula is only valid for a steady state, that is, when renal function is not changing rapidly.

The accurate measurement of creatinine clearance is difficult, especially in outpatients, as it is necessary to obtain a complete and accurately timed sample of urine. The usual collection time is 24 h, but patients may forget the time or forget to include some urine in the collection. Incontinent patients may find it impossible to make a urine collection. Patients have been known to add water or some other person’s urine to their own collection, hoping to gain the doctor’s approval for having been so prolific.

It may be more convenient and reliable to base the collection period on a patient’s normal habits (e.g. overnight). The time at which the bladder is emptied before retiring to bed is noted; any urine passed during the night is collected, as is the urine voided when the patient rises. The time is noted and a blood sample is taken that morning for the measurement of plasma creatinine. As long as the time over which the urine collection is made is known, and the collection is complete, any suitable time period can be used.

Creatinine is actively secreted by the renal tubules and, as a result, the creatinine clearance is higher than the true GFR. The difference is of little significance when the GFR is normal, but when the GFR is low (<10 mL/min), tubular secretion makes a major contribution to creatinine excretion and creatinine clearance significantly overestimates the GFR. The effect of creatinine breakdown in the gut also becomes significant when the GFR is very low. Certain drugs, including spironolactone, cimetidine, fenofibrate, trimethoprim and amiloride, decrease creatinine secretion and thus can reduce creatinine clearance. Lastly, in the calculation of creatinine clearance, two measurements of creatinine concentration and one of urine volume are required. Each of these has an inherent imprecision that can affect the accuracy of the overall result. Even in well-motivated subjects, studied under ideal conditions, the coefficient of variation of measurements of creatinine clearance can be as high as 10%, and it can be two or three times greater than this in ordinary patients.

Thus, although hitherto widely used, measurements of creatinine clearance are potentially unreliable and no longer recommended in routine practice. Alternative methods should be used if a reliable calculation of GFR is required, for example in the assessment of potential kidney donors, investigation of patients with minor abnormalities of renal function and calculation of the initial doses of potentially toxic drugs that are eliminated from the body by renal excretion.

There are two main alternative approaches to determining the GFR in clinical practice. These are to use exogenous markers of clearance or to derive an estimated GFR (eGFR) from the plasma creatinine concentration (see p. 67). GFR can be measured by measuring the disappearance from the blood of a test substance that is completely filtered by the glomeruli and neither secreted nor reabsorbed by the tubules, following a single injection. This approach has the advantage that a urine collection is not required. Suitable substances for this purpose include ⁵¹Cr-labelled EDTA (ethylenediaminetetra-acetic acid), ¹²⁵I-iothalamate (for which the decline in plasma radioactivity is monitored) and iohexol, a non-radioactive X-ray contrast medium that is simple to measure using high performance liquid chromatography. Typically, blood samples are taken at 2, 3 and 4 h after injection, although samples taken over a longer period may be required in renal impairment.

Plasma creatinine

Plasma creatinine concentration is the most reliable simple biochemical test of glomerular function. Ingestion of a meat-rich meal can increase plasma creatinine concentration by as much as 20 µmol/L for up to 10 h afterwards, so ideally blood samples should be collected after an overnight fast. Strenuous exercise also causes a transient, slight increase in plasma creatinine concentrations. Plasma creatinine concentration is related to muscle bulk and therefore a value of 120 µmol/L could be normal for an athletic young man but would suggest renal impairment in a thin, 70-year-old woman. Although muscle bulk tends to decline with age, so too does the GFR, and hence plasma creatinine concentrations remain fairly constant.

Some commonly used laboratory methods for the measurement of creatinine can suffer from interference, for example from bilirubin and ketones. The laboratory should be able to advise on whether this may be a problem in individual cases.

The reference range for plasma creatinine in the adult population is 60–120 µmol/L, but the day-to-day variation in an individual is much less than this range. Equation 4.1 indicates that plasma creatinine concentration is inversely related to the GFR. GFR can decrease by 50% before plasma creatinine concentration rises beyond the normal range; plasma creatinine concentration doubles for each further 50% fall in GFR. Consequently, a normal plasma creatinine does not necessarily imply normal renal function, although a raised creatinine does usually indicate impaired renal function (Fig. 4.3). Furthermore, a change in creatinine concentration, provided that it is outside the limits of normal biological and analytical variation, does suggest a change in GFR, even if both values are within the population reference range (see Case history 1.2).

Figure 4.3 Relationship between creatinine clearance and plasma creatinine concentration.

Changes in plasma creatinine concentration can occur, independently of renal function, owing to changes in muscle mass. Thus a decrease can occur as a result of starvation and in wasting diseases, immediately after surgery and in patients treated with corticosteroids; an increase can occur during refeeding. However, changes in creatinine concentration for these reasons rarely lead to diagnostic confusion.

In pregnancy, the GFR increases. This usually more than balances the effect of increased creatinine synthesis during pregnancy, and results in a decrease in plasma creatinine concentration.

Estimated GFR

An alternative to a formal measurement of creatinine clearance is to calculate an estimate of the clearance from the serum creatinine concentration. Various formulae have been derived for this purpose, including factors such as age, body weight, sex (creatinine production tends to be lower in women than men with the same body weight, because of their relatively smaller muscle mass) and racial origin. Several formulae have been derived from the Modification of Diet in Renal Disease (MDRD) study. The ‘four-variable’ formula is:

where [sCr] = serum creatinine concentration (µmol/L) and age is measured in years. This formula is for white males: the result should be multiplied by 0.742 for females and by 1.21 for African Caribbean people. (A calculator is available at: www.renal.org.)

A six-variable formula includes serum urea and albumin concentrations in addition. Because different laboratories may measure creatinine using different techniques, correction factors can be used to relate the calculated eGFR to the reference method for creatinine measurement, thus supposedly ensuring that all eGFR values are comparable. However, although the recommended formula for routine reporting in UK laboratories is the four-variable MDRD formula, it should be appreciated that it was derived from studies on patients with chronic kidney disease and may not be applicable to subjects with normal or near-normal renal function. Indeed, it is recommended that values of eGFR greater than 60 mL/min should be reported as ‘>60 mL/min’ and regarded as normal in the absence of clinical or laboratory evidence of renal disease (e.g. abnormalities on imaging, proteinuria or haematuria). Neither is it applicable in acute kidney disease, pregnancy, in conditions in which there is severe muscle wasting, oedematous conditions, amputees or in children. A major use of the MDRD eGFR is as a tool for screening for chronic kidney disease (CKD). UK guidelines recommend annual screening using eGFR and urinary protein in patients at risk of developing CKD, as summarized in Fig. 4.4. A more recent formula, the CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration), is based on pooled data from several studies and correlates better with measured GFR than the original MDRD formula, especially at values above 60 mL/min.

Figure 4.4 Early identification of chronic kidney disease.

Although now less widely used and not validated for use in screening for CKD, the Cockcroft–Gault formula, which is applicable over a wider range of GFRs, also provides an estimate of the creatinine clearance and hence GFR. Many drug dosing regimens, such as some of those used in cancer chemotherapy, are based on this, despite lack of adjustment for local creatinine calibration. The formula takes into account body weight in addition to sex, age and serum creatinine concentration:

Plasma urea

Urea is synthesized in the liver, primarily as a by-product of the deamination of amino acids. Its elimination in the urine represents the major route for nitrogen excretion. It is filtered from the blood by the glomeruli but significant tubular reabsorption occurs through passive diffusion.

Plasma urea concentration is a less reliable indicator of renal glomerular function than creatinine. Urea production is increased by a high protein intake, in catabolic states, and by the absorption of amino acids and peptides after gastrointestinal haemorrhage. Conversely, production is decreased in patients with a low protein intake and sometimes in patients with liver disease. Tubular reabsorption increases at low rates of urine flow (e.g. in fluid depletion), and this can cause increased plasma urea concentration even when renal function is normal.

Factors affecting the ratio of plasma urea to creatinine are summarized in Figure 4.5. Changes in plasma urea concentration are a feature of renal impairment, but it is important to consider possible extra-renal influences on urea concentrations before ascribing any changes to an alteration in renal function.

Figure 4.5 Causes of an abnormal plasma urea to creatinine ratio. ^aCauses of decreased creatinine synthesis; other conditions primarily affect urea concentration.

Urea diffuses readily across dialysis membranes and, during renal dialysis, a fall in plasma urea concentration is a poor guide to the efficacy of the process in removing other toxic substances from the blood.

Cystatin C

This low molecular weight peptide (13 kDa) is produced by all nucleated cells. It is cleared from the plasma by glomerular filtration and its plasma concentration reflects the GFR more accurately than creatinine. It is not influenced by gender or muscle mass but may be increased in malignancy, hyperthyroidism and by treatment with corticosteroids. Although not widely available in routine laboratories, measurement may have a role in the detection of early renal impairment in patients in whom creatinine is affected by unusual muscle bulk (e.g. body builders, teenage boys and small, elderly women).

Pre-renal acute kidney injury

This is caused by circulatory insufficiency, as may occur with severe haemorrhage, burns, fluid loss, cardiac failure, systemic sepsis or hypotension that leads to renal hypoperfusion and a decrease in GFR. This may in part be due directly to a fall in systemic blood pressure to below the level at which autoregulation can preserve the GFR, but can occur even if blood pressure is maintained, as this is achieved by sympathetic activation, which induces intense renal vasoconstriction. Initially, this results in a decrease in GFR with relative preservation of tubular function (allowing conservation of sodium and water, and hence ECF volume). However, if adequate perfusion is not rapidly restored, pre-renal uraemia may progress to intrinsic failure (‘acute tubular necrosis’). Older patients with CKD, diabetes or hypertension, or patients exposed to nephrotoxic drugs (e.g. aminoglycosides) or X-ray contrast agents are at particular risk of such progression. It may be possible to prevent this if renal perfusion can be restored before structural damage has occurred.

Pre-renal uraemia is essentially the result of a normal physiological response to hypovolaemia or a fall in blood pressure. Stimulation of the renin–angiotensin–aldosterone system and vasopressin secretion typically results in the production of a small volume of highly concentrated urine with a low sodium concentration (a fact that may be helpful in distinguishing between pre-renal and intrinsic AKI; Case history 4.1; Fig. 4.6). Renal tubular function is normal, but the decreased GFR results in the retention of substances normally excreted by filtration, such as urea and creatinine. Decreased excretion of hydrogen and potassium ions results in a tendency to metabolic acidosis and hyperkalaemia (the latter often being exacerbated by tissue damage).

Figure 4.6 Biochemical values in oliguria due to pre-renal and intrinsic kidney injury. Intermediate values occur in incipient intrinsic kidney injury.