The diagnostic and prognostic value of renal allograft biopsy.

We prospectively studied 89 patients to assess the diagnostic use of renal allograft biopsy in the first three months after transplantation. These biopsies were done in patients in whom diagnosis was not clear or clinical rejection was deemed to be severe. Clinical diagnosis at initial biopsy was compared with the morphological diagnosis. To determine if morphological data improved the prognostic usefulness of the clinical data, we performed multiple logistic regression relating clinical variables at initial biopsy and histological changes in the transplant to the outcome of 120 patients one year after biopsy. The clinical and morphological diagnosis differed in 41 of 89 patients (46%). Of 120 patients in the prognostic study, 35 returned to dialysis during the first year following transplantation. Using multiple logistic regression, a categorical variable that took into account both the serum creatinine and its rate of change before biopsy was the best clinical predictor of return to dialysis. Further increase in chi 2 occurred with type of donor, number of transfusions, and age. Using the clinical variables we produced an index, from 0 to 1 to predict outcome. Only 8 had index less than 0.2, of whom 7 returned to dialysis. The best morphological predictor of outcome was interstitial hemorrhage. Further increase in chi 2 was obtained with vascular endothelial proliferation, glomerular endothelial swelling, and glomerular necrosis. With an index derived from the morphological variables only 11 had index less than 0.2, of whom 9 returned to dialysis. Combining both clinical and morphological data, the best predictor of return to dialysis was interstitial hemorrhage, followed by creatinine, glomerular endothelial swelling, and type of donor. Using both clinical and morphological variables we produced another index to predict outcome. A group of 65 patients had index greater than 0.8, of whom 63 (94%) did not return to dialysis, and 18 patients had index less than 0.2, 17 of whom returned to dialysis. The remaining 12 patients in the dialysis group and 15 in the nondialysis group had indices greater than 0.2 less than 0.8. We conclude that a transplant biopsy yields important diagnostic and prognostic information. Unexpected diagnoses were made in 46% of cases. The addition of morphological data to the clinical data available at time of biopsy greatly improved the prediction of return to dialysis.