An essay into the various methods of assessing glomerular filtration.

                GFR =

Although very precise, this technique is rarely used because it requires a constant infusion of inulin, accurately timed urine and blood samples as well as complex assays for inulin makes this process prohibitively costly in clinical practice, despite being the best measurement in research medicine 3.

Another method for measuring glomerular filtration is similar to that of inulin, but instead uses radionuclides.  These have the advantage that they can be easily detected at very low concentrations using conventional counters.  Their major disadvantage is that the patients are subjected to radiation (albeit in very low doses) and that radiopharmaceuticals are more expensive than inulin 1.  The most commonly used radiopharmaceutical in measuring glomerular filtration is EDTA labelled with the gamma emitter 51Cr.  [51Cr] EDTA is filtered at the glomerulus and is not absorbed or excreted by the tubules.  It can therefore be accurately measured in both the blood and the urine, allowing a precise glomerular filtration rate to be calculated.  The clearance of [51Cr] EDTA by the glomerulus can be used to calculate the glomerular filtration rate from measurements of only a series of plasma concentration of [51Cr] EDTA measurements a few hours after an intravenous dose has been administered 3.  The isotopic activity of the samples can be plotted on a decay curve, which is related to GFR 4.  The advantage of this is that urine does not have to be collected, which is normally a major source of error, especially among children.

The most commonly measured endogenous substance used to assess glomerular function is creatinine, because of cheapness and ease of replication 3.  Creatinine is created by non-enzymatic dehydration of muscle 1 and is filtered at the glomerulus (although some is excreted by the tubules 7): its plasma concentration is therefore dependent on muscle mass and glomerular filtration rate.  When measuring the clearance of creatinine, blood and urine samples are collected over 24-hours.  The urine is then analysed for volume and creatinine concentration, and the blood is analysed for serum.  Because creatinine is secreted by the tubules, glomerular filtration rate is overestimated when the kidney is normal, but in advanced renal failure less creatinine is filtered, so the amount secreted is proportionally greater, and thus the GFR is greatly over-estimated 4.  There are also problems obtaining accurately timed urine samples, accurate creatinine measurements, and variation in generation rate of creatinine.  A change in muscle mass will lead to a change in creatinine production, which could be mistaken for a change in glomerular filtration rate, as it takes up to 3 weeks for urinary excretion of creatinine to equilibrate.  Stewed meat, such as ‘goulash’, has also been implicated in a doubling of plasma creatinine 1, which will result in a misleading measurement of glomerular filtration.

A simplification of creatinine measurement is to use only a serum creatinine measurement to estimated glomerular function.  Serum creatinine is proportional to muscle mass and glomerular filtration rate, so alterations in glomerular filtration rate will be reflected in the serum creatinine level 1.  Various formulae have been developed to calculate GFR from the plasma creatinine: these must take into account age, sex and weight to allow for differing muscle masses between individuals.  One useful formula which can approximately calculate the creatinine clearance is the Cockcroft-Gault formula. 2

(140age)×body weight (kg)×K

                            serum creatinine (µmol/l)

K (constant) is 1.25 for men and 1.03 for women.

The major source of error using this method is that although creatinine levels stay constant throughout life, muscle mass diminishes and the number of functioning nephrons falls: these factors should balance each other, as age increases the possibility of error increases, resulting in normal creatinine levels not necessarily indicating normal glomerular function.

The measurement of blood urea is not a good indication of glomerular filtration, as it is affected by protein catabolism urea can be passively reabsorbed by the nephrons if the flow rate is low 3.  It is therefore not used in clinical practice.

The kidney provides an important area for the breakdown of peptides and small proteins, which occurs mainly by filtration at the glomerulus followed by pinocytotic reabsorption in the tubule.  Therefore any accumulation of peptides is indicative of declining renal function.  The two main micro-proteins researched are β2-microglobulin and cystatin A 1.  The advantages of testing for these endogenous substances are that their production is unaffected by age, sex or muscle mass.  The disadvantage of measuring β2-microglobulin is that lymphatic tumours and rheumatoid arthritis can raise its plasma concentration; its measurement is also difficult and expensive.  However cystatin A is not raised by malignancies or inflammatory states 1, and with the introduction of its measurement using automated immunoturbidimetric methods (which have been shown to be as accurate as inulin clearance), it is very likely the measurement of cystatin A will become commonplace in clinical practice.

A final new technique for measuring glomerular filtration rate employs a gamma camera to obtain images of the kidney as it filters radioactive [99Tcm] DTPA 5.  The GFR can be calculated using time-lapsed images, and results within 3% of inulin are impressive and warrant further research 6.

In ordinary clinical practice, at present the measurement of plasma creatinine is an adequate method of estimating glomerular function provided that urine flow is normal.  If not, then glomerular filtration should be measured using a constant intra-venous infusion of [51Cr] EDTA or inulin with blood samples.  These methods are also essential for clinical trials.  The new methods to estimate glomerular function, especially cystatin A and [99Tcm] DTPA, seem to have less margin for error and appear to be as accurate as inulin clearance.

---

---