Computer-assisted evaluation enhances the quantification of interstitial fibrosis in renal implantation biopsies, measures differences between frozen and paraffin sections, and predicts delayed graft function.

BACKGROUND: (Pre-)Implantation biopsies provide important data on the quality of donor kidneys. Interstitial fibrosis, as a known predictor for kidney disease progression, is an essential feature of this evaluation. However, the assessment of frozen sections of implantation biopsies is challenging and can result in the disposal of candidate organs. We sought to apply digital image analysis (DIA) to quantify the differences between frozen and paraffin sections when evaluating interstitial fibrosis, identify factors that influence these variations and test the predictive value of the computerised measures. METHODS: We quantified the differences between frozen and paraffin sections in the same biopsy samples by measuring Sirius red-stained interstitial areas (SRIA) in DIA. We compared them to the original reports, and retrospectively correlated our findings to clinical data, graft function and outcome in 73 patients. RESULTS: Frozen sections display a broader interstitial area than paraffin sections, in some cases up to one-third more (mean difference + 7.8%, range - 7 to 29%). No donor-related factors (age or gender, cold ischemia time, or non-heart-beating donor) influenced significantly this difference. Compared to the original assessment of frozen vs paraffin sections in optical microscopy, the DIA of interstitial fibrosis shows a higher consistency (ICC 0.69). Our approach further allows to distinguish SRIA in paraffin sections as an independent predictor for delayed graft function (OR = 1.1; p = 0.028). CONCLUSIONS: DIA is superior to and more consistent than routine optic microscopy for interstitial fibrosis evaluation. This method could improve implantation biopsy diagnostics and help to reduce disposal of organs.