Purpose: Inflammation in regions of tubular atrophy (IATR) is a risk factor for late graft loss. However, IATR is associated with inflammation in non-atrophic regions of the biopsy and with interstitial fibrosis. Our objective was to determine whether IATR adds any risk beyond degree of fibrosis, inflammation in non-atrophic regions, creatinine at time of biopsy, and time from transplant to biopsy.

Methods: The study population consisted of 480 patients with late onset (median time from transplant to biopsy 6.1 years) graft dysfunction (defined as a rise in serum creatinine of 25% over baseline or new onset proteinuria) . Fibrosis scores were measured in three ways: Model 1-using the Banff ci score (0-3); Model 2 -using percent fibrosis assessed by quantitative image analysis; and Model 3-using a pathologist estimate of fibrosis. Cox proportional hazards models were used to estimate risk of graft loss using IATR scores (0-3), inflammation in non-atrophic regions of the biopsy (Banff I score), fibrosis scores (Models 1-3), creatinine at time of biopsy, and time from transplant to biopsy. Assessments were made by a central pathologist blinded to specimen identity. The outcome was time to death censored graft failure from time of biopsy.

Results: In all 3 models, Banff I scores and time to biopsy were not significant predictors of graft failure. In all 3 models, fibrosis and serum creatinine at time of biopsy strongly correlated with graft outcome. For Model 1, a Banff ci score of 1 versus 0 was a borderline predictor of graft failure (hazard ratio 1.78, CI 1.00 – 3.17, p=0.051). Akaike Information Criterion (AIC) analysis indicated that Model 2 (%fibrosis determined by quantitative analysis) was the best model predictive of graft failure. In this model, each additional percentage point of fibrosis increased the hazard ratio by 2.2% (1.004 – 1.040, p=0.0177). In all models, IATR remained a strong predictor of adverse graft outcome.

Conclusions: Banff i scores did not predict adverse graft outcome, possibly because these patients were treated for acute rejection episodes. IATR is a strong predictor of adverse graft outcome, even when adjusted for inflammation in non-atrophic areas and for fibrosis. We conclude that IATR should be assessed in patients with late graft dysfunction.

« Back to 2015 American Transplant Congress