Predictors of Late Kidney Dysfunction Post-Liver Transplantation E. Velidedeoglu, M.D. Crawford, N.M. Desai, L. Campos, P.L. Abt, J.W. Markmann, K.C. Mange, R. Bloom, K.M. Olthoff, A. Shaked, and J.F. Markmann R ENAL DYSFUNCTION is common among patients awaiting liver transplantation and represents an im- portant determinant of posttransplant outcome. 1 Less clear, however, is the relationship between acute renal dysfunc- tion (ARD) in the perioperative period and subsequent development of chronic renal dysfunction (CRD) and post- transplant long-term survival. Moreover, the relationship of ARD and CRD post-liver transplant to other predisposing factors, such as the nephrotoxic effects of immunosuppres- sants and various other drugs, is poorly defined. Although there is no general consensus on the precise definition of ARD and CRD, the incidences of both complications are quite high after liver transplantation, namely 80% to 90% and almost 80% respectively, at 3 to 5 years after liver transplant. 2,3 The incidence of end-stage renal disease (ESRD) is 1.4% at 4 years and almost 10% at 13 years after the liver transplant. 1 The current retrospective study was conducted to elucidate possible risk factors that contribute significantly to the development of posttransplant CRD as a means to identify the subpopulation of recipients who are at high risk for the development of CRD and ultimately ESRD for appropriate interventions such as early conver- sion to nonnephrotoxic immunosuppressions. PATIENTS AND METHODS We retrospectively reviewed the outcomes of 295 adult orthotopic liver transplants performed between 1/1/98 and 12/31/00. After exclud- ing patients who received other organ transplants simultaneouly, obtained organs from living donors, were retransplants, experienced primary nonfunctions and all deaths during the study and follow-up period, the final study population consisted of 189 patients who were alive with good liver function at the completion of the follow-up period (1/1/02). The reason for the exclusion of patients who died was that these patients frequently developed ARD during the terminal stage of illness regardless of their prior renal function. Thus, inclusion of these patients may lead to misleading results. The 189 orthotopic liver transplant (OLT) recipients were analyzed for the impact on renal function of a variety of parame- ters: patient demographics, type of liver disease, preoperative renal function, hepatorenal syndrome, cumulative calcineurin inhibitor dosages received and levels reached during the first postoperative week, preoperative and postoperative diabetes requiring insulin treatment, and kidney function at various time points after the transplant. To assess long-term kidney function, last follow-up lab data were used for patient (mean follow-up for the cohort was 2.11 years; minimum follow-up was 9 months). For the purposes of this analysis, renal dysfunction was defined as a serum creatinine (Cr)  2 mg/dL. The SPSS statistical software package was used for chi-square analysis of variables in the data set. RESULTS Mean recipient age was 49.6 years (20 to 74), 65% were men, and all patients received calcineurin inhibitors as primary immunosuppressive therapy including 160 patients (85%) treated with tacrolimus and 29 (15%) with cyclospor- ine. The incidence of preoperative ARD was 4.7% (n = 9). The incidence of Cr  2 mg/dL during the first week was 38% (n = 72), with a mean peak Cr of 3.4 mg/dL in this subgroup. The incidence decreased to 11% (n = 22) by day 14, 7% (n = 15) by day 30, and 3% (n = 7) by day 90. Long-term renal dysfunction occurred in 6.3% (n = 12), with only one patient progressing to ESRD and requiring hemodialysis (0.5%). Among of the 12 patients who devel- oped CRD, 10 (84%) displayed peak values of serum Cr  2 mg/dL during the first week. However, 8 of these 10 patients had normalized their Cr by day 90 (six patients Cr  1.5 mg/dL and two patients Cr  2 mg/dL). In our series, the overall incidence of diabetes preoperatively and post- operatively was 22.8% (n = 43) and 55.0% (n = 104), respectively. Insulin treatment was required in 11.6% (n = 22) and 32.3% (n = 61), respectively. Among the patients who had ARD (n = 72) posttransplant, the pre- and posttransplant incidence of diabetes were 28% (n = 20) and 60% (n = 43). Of these 29 patients who had both ARD and diabetes requiring insulin treatment, six patients (21%) developed CRD. The association between ARD and preop and postop diabetes did not reach statistical significance (P = .188 and .309). Of the 12 patients who developed CRD, six had preoperative diabetes five required insulin treat- ment) and two more patients developed diabetes requiring insulin therapy after the transplant, making the incidence of diabetes 58% (n = 7) in this subpopulation. Late kidney dysfunction was correlated with three perioperative vari- ables by chi-square analysis: peak Cr  2 mg/dL during the first week after the transplant (P = .001), preoperative From the Department of Surgery, Hospital University of Penn- sylvania, Philadelphia, Pennsylvania, USA. Address reprint requests to E. Velidedeoglu, Department of Surgery, Hospital University of Pennsylvania, 3400 Spruce St, 4th Floor Silverstein, Philadelphia, PA 19104. © 2002 by Elsevier Science Inc. 0041-1345/02/$–see front matter 360 Park Avenue South, New York, NY 10010-1710 PII S0041-1345(02)03627-8 Transplantation Proceedings, 34, 3315–3316 (2002) 3315