Background: Various methods have been used to determine mean glomerular volume, some requiring measurement of over 30 glomerular profiles for a satisfactory estimate. Needle biopsies are useful diagnostically, but if small, provide insufficient tissue for the use of such methods.
Methods: We performed glomerular volume measurements on renal biopsies from 10 normotensive, non-uraemic patients with Type 1 diabetes. Sections were taken at 10 microm intervals through 10 glomeruli per biopsy and points landing on glomerular tuft counted under light microscopy. Volume was calculated from the measured cross-sectional area and known section thickness using the Cavalieri principle.
Results: Estimating the volume of 10 glomeruli per biopsy gave an overall mean glomerular volume of 4.21x10(6) microm(3) and standard deviation between patient means 1.23x10(6) microm(3.) Using a sample size of five glomeruli per biopsy only increased the standard deviation between patient mean values by 3%. Using sections taken at 20 microm intervals made little difference to the mean glomerular volume and standard deviation estimates (MGV 4.20x10(6) microm(3)+/-1.24). Further increases in the sectioning interval resulted in an appreciable increase in the variance of the estimate.
Conclusions: The results suggest that a satisfactory estimate of mean glomerular volume can be obtained from a sample size of five glomeruli per biopsy using a sectioning interval of 20 microm. This represents a great saving in analysis time and effort, making widespread use of this method of glomerular volume measurement in renal disease more practicable, in both research and clinical settings.