Measurements of cortical interstitium in biopsies from human kidney grafts: how representative and how reproducible?

Background. The aim of the study was to evaluate how well a biopsy from one region of a human renal allograft represented biopsies from other regions regarding the renal interstitial tissue assessed by stereology. Furthermore, we wanted to evaluate the reproducibility of the measurements. Methods. Punch biopsies (3 mm) from six regions in each kidney were obtained from seven explanted renal grafts with varying degrees of clinically diagnosed chronic and acute rejection. One kidney, removed for a minor pelvic tumour, served as reference material. Using point counting on PAS-stained sections, the volume fraction of the interstitial tissue per glomerular cortex V-V(interstitium/cortex) was estimated. From each kidney, two of the six biopsies were re-evaluated by the same observer. Results. V-V(interstitium, cortex) varied from 0.25 to 0.78 between the explanted kidneys vs 0.26 in the reference kidney. Variations within the kidneys were low, expressed by standard deviations (SD) of between 0.04 and 0.06, and coefficients of variation (CV) between 0.06 and 0.22. The SD estimated from repeated measurements was 0.04 and CV was 0.07. Conclusions. Biopsies from one region of the kidney were found to be representative for estimates of interstitial tissue in explanted human kidney grafts, and the degree of reproducibility was high when using point counting, as in the present study.