Validation Of A Predictive Model For Chronic Renal Dysfunction Following Liver Transplantation

Background and Aims: Chronic renal dysfunction (CRD) is a frequent complication after liver transplantation (LT).We prospectively analyzed the incidence of CRD after LT and we looked for possible predictors which may help to identify these patients during follow up.The predictive model was validated in an independent population of transplanted patients at the University of Valencia.Methods: One-hundred-sixty-one patients (128 males, mean age 54±9 years, follow-up 46±30 months undergoing LT in the period January 2000 to August 2009 at the University "La Sapienza" of Rome were included in the study to build the predictive model.Survival or follow-up <3 months; diagnosis of organic renal failure; no compliance to periodic follow-up were exclusion criteria.The diagnosis of CRD was based on serum creatinine ≥1.5 mg/dl in two subsequent controls.Serum Creatinine (SC), diabetes, arterial hypertension, acute allograft rejection, blood level of immunosuppressive therapy, episodes of severe infections, development of cirrhosis were taken into consideration sequentially at 12-month intervals.These variables were used to build a time-dependent Cox's regression model to assess possible predictors of the occurrence of renal failure within 12 months of every assessment.One hundred-fifty patients transplanted at the University La Fe of Valencia in the same period were used for validation.Result: Cumulative incidence of CRD was 17.3% (95% CI: 12-24) at one year, 22.3% (95%Cl: 16-30), 27.6% (95% CI: 20-36), 34.4% (95% CI: 26-44) at 2, 3 and 5 years after LT respectively.SC, arterial hypertension, episodes of severe infection, development of cirrhosis, resulted to be significant and independent predictors of the onset of CRD within 12 months.Validation was performed dividing the validation sample in a group with high and low risk of CRD according to the data obtained in the training sample.Afterwards, expected probability curves were calculated for the groups with low and high risk of CRD, and were compared with the Kaplan-Meier estimates of the observed outcome for the same patients.A reasonable concordance was observed (see figure 1).Figure 1.Conclusions: Easy available data may help to calculate the probability to develop CRD after liver transplantation at yearly intervals.